NC0000272_SewerGeneratedColorInfo_20031015

BLUE RIDGE PAPER PRODUCTS INC.
Xc: Betsy Bicknell
Dan Bodien
Paul Davis
Norm Liebergott
Neil McCubbin
Dave McKinney
Karrie-Jo Shell
Ahmar Siddiqui
+korres .:estall, _
➢ d
2J �i Off 16 2003
i ff
-_�f pil:I I"!SEC110?I

BLUE RIDGE
PAPER PRODUCTS INC.
October 15, 2003
Mr. Don Anderson
Chief, Commodities Branch
Engineering and Analysis Division
USEPA Headquarters (4303)
Room 195A, East Tower
401 Main Street
Washington, D.C. 20460
RE: Canton Mill sewer generated color information request
Dear Mr. Anderson:
Enclosed is the information you requested for the Technology Review Workgroup in
reference to sewer generated color at the Canton Mill of Blue Ridge Paper Products Inc.
Please do not hesitate to contact Bob Williams at [phone] or myself at(828) 646-
6749 if you have any questions about the attached information.
Sincerely,
-WMOV,61dw
Melanie Gardner
Adv. Environmental Engineer
Blue Ridge Paper Products Inc.
Attachment
175 Main Street • P.O. Box 4000
Canton, North Carolina 28716
Phone: [phone] - Fax: [phone]
Raising Your F-xpectations

Section 1
Mill Color Balance: 1999 — 2003
Annual Average Color (Ibs/day)
Primary Influent Secondary Effluent Sewer Generated Color
1999 54,437 41,086 3,740'
2000 57,558 43,355 10,732'
2001 59,163 43,424 14,0901
Jan-02-Sep-02 54,732 40,556 2,4852
Nov-02-Apr-03 61,664 49,027 20,6912
May-03-Aug-03 54,728 42,670 9,4912
There were quantifiable events in the 1999 - 2001 periods that contributed to
these averages, unlike the 2002 — 2003 elevated color period when measurable
events were not the major contributors.
2 A thorough investigation of in-mill color sources was conducted and one
unmeasured color source was pinpointed. The hardwood screen reject stream is
sewered at the rate of 100 — 150 gpm. The average color contribution from this
stream ranged from 3,000— 5,000 Ibs/day historically and is currently in the
2,000 — 3,000 Ibs/day range due to process changes that reduced the color. This
contribution has been subtracted from the average unaccounted color since
2002. During the November 2002 —April 2002 period, it is clear that the
Unaccounted color had increased significantly despite accounting for the
hardwood screen rejects.
The attached color pies (Figures 1 — 6) demonstrate the significant increase in
unaccounted (sewer generated) color that occurred beginning in November and
lasting through April 2003.
Information Submittal to EPA
Blue Ridge Paper Products Inc.
Canton, North Carolina
NPDES Permit#0000272
October 14, 2003

1999 Average Measured Color (Ibs/day) in Mill Sewers as a Percentage of
Primary Influent Color
PI Color = 54,437 Ibs/day
SE Color = 41,086 Ibs/day
4% 1%
4%
7%
27% © 1&2 FL's D1 + Pine D2: 14,819 ppd-27%
t7 58- Recovery, BLO,CRP: 13,320 ppd -24%
10% 0 3A-No. 1/2 Eo, No.2FL BSW, 02 Delig: 12,316 ppd -23%
O2B- Digesters, No. 1 FL: 5,666 ppd - 10%
m Unaccounted: 3,740 ppd -7%
■ PM's- 11 & 12: 2,090 ppd -4%
❑Contaminated Condensate:2,039 ppd-4%
23%
24% ❑Combined Condensate: 547 ppd - 1%
Blue Ridge Paper Products Inc.
Figure 1 Canton Mill MSG10/15/2003

2000 Average Measured Color (Ibs/day) in Mill Sewers as a Percentage of
Primary Influent Color
PI Color = 57,558 Ibs/day
SE Color = 43,355 Ibs/day
3% 1%
5%
6%
27%
IE 1&2 FL's D1 + Pine D2: 15,671 ppd -27%
0 5B- Recovery, BLO,CRP: 11,500 ppd -20%
© 3A- No. 1/2 Eo, No.2FL BSW, 02 Delig: 11,021 ppd- 19%
19% GO Unaccounted: 10,732 ppd - 19%
IE2B- Digesters, No. 1 FL: 3,687 ppd - 6%
N PM's- 11 & 12: 2,666 ppd -5%
©Contaminated Condensate: 1,976 ppd -3%
❑Combined Condensate: 304 ppd - <1%
20%
19%
Blue Ridge Paper Products Inc.
Figure 2 Canton Mill MSG1 011 5/2 003

2001 Average Measured Color (Ibs/day) in Mill Sewers as a Percentage of
Primary Influent Color
PI Color = 59,163 Ibs/day
SE Color = 43,424 Ibs/day
7%
r
O2B- Digesters, No. 1 FL, 4219 Ibs/day
18%
O 3A- No. 112 Eo, No.2FL BSW, 02 Delig, 10,839
Ibs/day
0 PM's- 11 & 12, 2,284 Ibs/day
0% 0 56- Recovery, BLO,CRP, 11,934 lbs/day
5/° u 4% W 1&2 FL's D1 + Pine D2, 12,423 Ibs/day
1� 2v I]Contaminated Condensate, 3,144lbs/day
OCombined Condensate, 229lbs/day
tr ❑Unaccounted, 14,090lbs/day
Blue Ridge Paper Products Inc.
Figure 3 Canton Mill [Date]

January- September 2002 Average Sewer Area Color as a Percentage
of Primary Influent Color (Ibs/day)
Primary Influent for Period = 54,732 Ibs/day
Secondary Effluent for Period = 40,556 Ibs/day
4.2% 0.9%
4.3% ® 513- Recovery, BLO,CRP (13,358 Ibs/day) 25%
4.6%
M 3A- No. 1/2 Eo, No.2FL BSW (11,035 Ibs/day)20%
10.1 IM 1&2 FL's D1 + Pine D2 (10,721 Ibs/day) 20%
O2B- Digesters, No. 1 FL(6,131 Ibs/day) 11%
HIScreen Rejects (5,487 Ibs/day) 10%
11.3% O Unaccounted (2,485 Ibs/day) 5%
10 PM's- 11 & 12(2,361 Ibs/day) 4%
20.3%
WContaminated Condensate (2,281 Ibs/day) 4%
19.7% ❑Combined Condensate (5131bs/day) 1%
BRPP Inc.
Figure 4 Canton, NC 10/15/2003

May 2003 - August 2003 Average Sewer Area Color as a Percentage of
Primary Influent Color (Ibs/day)
Period Average SE Color = 42,670 Ibs/day
Period Average PI Color = 54,728 Ibs/day
4% 1%
4%
6% 25%
M 513- Recovery, BLO,CRP (12,119 Ibs/day-25%)
m
® 3A- No. 1/2 Eo, No.2FL BSW (13,097 Ibs/day-22%)
6%
i Unaccounted (9,491 Ibs/day- 17%)
M 1&2 FL's D1 +Pine D2 (8,687 lbs/day- 15%)
OScreen Rejects for 2003 (3,529 Ibs/day- 6%)
❑Contaminated Condensate (1,5131bs/day- 6%)
15%
192B- Digesters, No. 1 FL(2,419 Ibs/day-4%)
0 PM's- 11 & 12 (1,269 Ibs/day-4%)
22%
■Combined Condensate (204lbs/day< 1%)
17%
BRPP Inc.
Figure 6 Canton Mill 10/15/2003

Section 2
Color Challenges in 2002 — 2003
Sewer generated color began increasing in November 2002. It was not readily
clear what the underlying factors were.
Despite around-the-clock sewer monitoring and 2-hour color testing, the cause of
this color was not found. Note the elevated 2-hour color results from the 2-hour
Color Test Results trend (Figure 7a, 7b shows a more recent, lower color 2-hour
Test Result trend). Elevated 2-hour color results would prompt operations
personnel to closely investigate potential color losses in their areas, but there
were consistently no sources of brown color found in the sewers. Also, there
was consistently no evidence of brown color losses on the mill's process control
monitoring systems (i.e., no tank overflows, no major in-mill sewer conductivity
spikes, etc.). Figure 8 shows the Plant Information system's conductivity display.
Note that low/no color chloride solutions also cause conductivity spikes,
specifically in 5A sewer.
The attached "Color Loadings to Mill Sewer Areas" reports (Figures 9 through 15)
show the levels of unaccounted color that were present during a representative
time in this period. Compared to the January 2002 — September 2002 "good"
color period (when unaccounted color averaged only 5%), it is clear that a
change within the mill sewers had taken place. In the past, the Color Loadings
report had indicated which sewer areas were elevated or out of control, but
during the November 2002 —April 2003 period, unaccounted color was
consistently elevated and the cause was not evident. Other examples of in-mill
sewer monitoring are attached including the mill sewer system with fiber losses,
color, flow and sampler locations shown (Figure 16) and the instantaneous sewer
conductivities with Primary Influent and Secondary Effluent flow and pH (Figure
17).
Following repeated episodes of searching for "mystery" color losses, a color team
was formed to investigate potential contributors to the sewer generated color.
Representatives from all areas of the mill participated in the color team's efforts.
Operational changes within each area during the period of elevated sewer
generated color were discussed. It was still not clear why the amount of sewer
generated color had significantly increased but there were several parameters
that were operating outside historical levels of variability.
Information Submittal to EPA
Blue Ridge Paper Products Inc.
Canton, North Carolina
NPDES Permit#0000272
October 14, 2003

A statistical approach was taken to identify which, if any, of the operational
parameters were significantly contributing to the increased color. Multiple
Analysis of Variance (ANOVA) Factorials were performed and the results of
these studies are attached to this section (summary of two ANOVAs shown in
Figures 18a and 18b). Separate analyses were done using the Secondary
Effluent as a dependent variable and also using unaccounted color as a
dependent variable. The ANOVA Factorial determines the significance of the
effects of the independent variables (i.e., the operational parameters discussed
by the color team) on the dependent variables.
Surprisingly, there were several main (one significant parameter) and interaction
(significant combinations of parameters) effects contributing to both elevated
Secondary Effluent color and unaccounted color during the high color period
(November 2002 —April 2003). It is believed that a type of domino effect was
seen in the Pulp mill and Recovery areas from certain operational problems, and
this caused the complex interaction effects.
Periods of good color performance were also analyzed and very few effects
showed up as being significant, further proving that the combination of certain
operational variables being elevated can intensify sewer generated color.
The results of the ANOVAs were used to create a Daily Color Monitoring Report,
which is also attached to this section (Figure 19). Part of this report is a daily
Secondary Effluent trend by month showing month-to-date performance with
monthly targets to achieve 40,000—42,000 Ibs/day averages.
Information Submittal to EPA
Blue Ridge Paper Products Inc.
Canton, North Carolina
NPDES Permit#0000272
October 14, 2003

` 1 i S N a
1
2-Hr Color Test Results MA: COLOR . L
i 1000. 0 - — -- - - - ❑ 717. OG
F'PPI
y ��J
0 .
2-Hr Projected Daily Color - - -� -- 1-14eCt3l0R . 0
#JDayF
120000 . � I
25-Dec-02 00 :00 00 29-Dec-02 04 : 13: 58 12hr/div 31-Dec-02 00 : 00: 00
Moue the cur,-,ur tc, a new pooiliufl

2-H!r Color lest, Imesult s I urn:CO Off . L
1000. 0 O(i
I
0 .
2_-Hr Projecte6 Daily Color hh crPLLOP . 0
120000 .
I~#lDay
LI
20000 .
11-OGL-,On 010 : 00: 00 13--Oct-OH 00 : 00: 00 121ir/cliv 18-0t.L-O'N 00 : 00: 00
chiii)(Ii, 111r1 L !i!r? sC :: lr' 0 `L !' ii lei ' E.

FTime Tag i ! t Display
0 Fi��s:;��gr-;; uin• Conti ;;��uer Con�lii�� l � ���'1 ;� Tr�� n�l•�; t;i- 0�� ( - �i: 1 '� : ,il : `>1
Figure 8
WIP IMFLUEFII` �ItdO. 2 �EWE� AREf� IiEt1t�POf�ATt1R SLIMF�S
FIBE�LFf7ES I - - -- -
'
STRIF'PE� SUI�iP - - I�ECOUEf:Y BAILER SEWERS - t�0. 6 SEWFh �H l
L
1 - Ot: t - b3 1 '; ' a! : '; 1 61TP IIIFLUEIIT 1a- 0�: 1 - U1s 1 'i , si : '�1

Color Loadings to Mill Sewer Areas 25-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 54793 Lbs/day 65705 53400 81963
Pri.Infl.New NCASI* 44505 Lbs/day
Pri Inf Turbidity 32 .00 18 .69 12 . 71
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78,609** Lbs/day
Secondary Effluent 43611 Lbs/day 47737 39472 50032
#1 Sewer 80 Lbs/day 2820
#2 Sewer 60 PPM 847
2B Sewer 500 Lbs/day 6965 4501 7837
3A Sewer 7210 Lbs/day 12838 12309 20228
#4 Sewer 172 PPM 342
Contaminated Condensate Sewer Values
Concentration 145.00 mg/L
Flow 766 .34 GPM
Color 1333.43 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 19 . 00 mg/L
Flow 406. 92 GPM
Color 170 . 91 Lbs/day 1186 1216 2251
5B Sewer 6090 Lbs/day 10508 7434 13581
CRP Color 4828. 73 Lbs/day 4965
6A Sewer 4470 Lbs/day 16736 13671 23053
6A Sewer pH 3 .90
Pine Bleach 10430 Lbs/day 9627 10119 18011
19 . 86 Lbs/ton pulp 16 .32 17.27 31.44
Pine D1 760 Lbs/day 5539 6420 12538
Pine Eo 9610 Lbs/day 2687 2658 4254
Pine D2 60 Lbs/day 871
Hardwood Bleach 11900 Lbs/day 12753 12439 21190
16.45 Lbs/ton pulp 17.25 14 .2 30 . 1
Hardwood D1 6330 Lbs/day 5905 5292 13435
Hardwood Eo 5570 Lbs/day 6848 5875 11815
BFR Closure 83. 6 %
Unaccounted Color 36211.1 Lbs/day
Sewer Total 18582.0 Lbs/day
* New NCASI color test method which accounts for turbidity.
. **The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 9

Color Loadings to Mill Sewer Areas 26-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 58045 Lbs/day 65705 53400 81963
Pri. Infl.New NCASI* 51741 Lbs/day
Pri Inf Turbidity 11 .40 18.69 12 . 83
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day
Secondary Effluent 37393 Lbs/day 47737 39472 50032
#1 Sewer 430 Lbs/day 2820
#2 Sewer 50 PPM 847
2B Sewer 5060 Lbs/day 6965 4501 7837
3A Sewer 3510 Lbs/day 12838 12309 20228
#4 Sewer 105 PPM 342
Contaminated Condensate Sewer Values
Concentration 124 . 00 mg/L
Flow 819 .03 GPM
Color 1218 .72 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 22 .00 mg/L
Flow 443 .87 GPM
Color 101.20 Lbs/day 1186 1216 2251
5B Sewer 8510 Lbs/day 10508 7434 13581
CRP Color 4036.53 Lbs/day 4965
6A Sewer 7890 Lbs/day 16736 13671 23053
6A Sewer pH 4 .20
Pine Bleach 10170 Lbs/day 9627 10119 18011
19 .37 Lbs/ton pulp 16 .32 17 .27 31.44
Pine D1 3580 Lbs/day 5539 6420 12538
Pine Eo 6560 Lbs/day 2687 2658 4254
Pine D2 30 Lbs/day 871
Hardwood Bleach 10010 Lbs/day 12753 12439 21190
12 . 61 Lbs/ton pulp 17 .25 14 .2 30 . 1
Hardwood Dl 5800 Lbs/day 5905 5292 13435
Hardwood Eo 4210 Lbs/day 6848 5875 11815
BFR Closure 70. 6 %
Unaccounted Color 32490 .5 Lbs/day
Sewer Total 25555 .0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 10

Color Loadings to Mill Sewer Areas 27-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 53960 Lbs/day 65705 53400 81963
Pri.Infl.New NCASI* 46495 Lbs/day
Pri Inf Turbidity 12 . 70 18. 69 13 . 05
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day -
Secondary Effluent 42229 Lbs/day 47737 39472 50032
#1 Sewer 710 Lbs/day 2820
#2 Sewer 128 PPM 847
2B Sewer 1900 Lbs/day 6965 4501 7837
3A Sewer 8570 Lbs/day 12838 12309 20228
#4 Sewer 251 PPM 342
Contaminated Condensate Sewer Values
Concentration 301. 00 mg/L
Flow 804 .32 GPM
Color 2905 . 19 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 49 . 00 mg/L
Flow 412 . 14 GPM
Color 108.81 Lbs/day 1186 1216 2251
5B Sewer 11000 Lbs/day 10508 7434 13581
CRP Color 3581.24 Lbs/day 4965
6A Sewer 6110 Lbs/day 16736 13671 23053
6A Sewer pH 3. 60
Pine Bleach 11470 Lbs/day 9627 10119 18011
18.95 Lbs/ton pulp 16.32 17.27 31.44
Pine Dl 1680 Lbs/day 5539 6420 12538
Pine Eo 9710 Lbs/day 2687 2658 4254
Pine D2 80 Lbs/day 871
Hardwood Bleach 14170 Lbs/day 12753 12439 21190
16.99 Lbs/ton pulp 17.25 14 .2 30 . 1
Hardwood D1 6760 Lbs/day 5905 5292 13435
Hardwood Eo 7410 Lbs/day 6848 5875 11815
BFR Closure 66.7 %
Unaccounted Color 25290 .5 Lbs/day
Sewer Total 28669 . 0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 11

Color Loadings to Mill Sewer Areas 28-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 81548 Lbs/day 65705 53400 81963
Pri.Infl.New NCASI* 74803 Lbs/day
Pri Inf Turbidity 13 . 00 18. 69 12 .99
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day
Secondary Effluent 44759 Lbs/day 47737 39472 50032
#1 Sewer 840 Lbs/day 2820
#2 Sewer 218 PPM 847
2B Sewer 1910 Lbs/day 6965 4501 7837
3A Sewer 19310 Lbs/day 12838 12309 20228
#4 Sewer 298 PPM 342
Contaminated Condensate Sewer Values
Concentration 420 .00 mg/L
Flow 435 .26 GPM
Color 2193 .70 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 62 .00 mg/L
Flow 420 .82 GPM
Color 247 .44 Lbs/day 1186 1216 2251
5B Sewer 10630 Lbs/day 10508 7434 13581
CRP Color 3501.98 Lbs/day 4965
6A Sewer 7340 Lbs/day 16736 13671 23053
6A Sewer pH 3 .40
Pine Bleach 9270 Lbs/day 9627 10119 18011
37 .20 Lbs/ton pulp 16 .32 17 .27 31 .44
Pine Dl 350 Lbs/day 5539 6420 12538
Pine Eo 8880 Lbs/day 2687 2658 4254
Pine D2 40 Lbs/day 871
Hardwood Bleach 16080 Lbs/day 12753 12439 21190
19. 75 Lbs/ton pulp 17 .25 14 .2 30 . 1
Hardwood D1 6980 Lbs/day 5905 5292 13435
Hardwood Eo 9100 Lbs/day 6848 5875 11815
BFR Closure 61.6 %
Unaccounted Color 41001.6 Lbs/day
Sewer Total 40546.0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 12

Color Loadings to Mill Sewer Areas 29-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 75489 Lbs/day 65705 53400 81963
Pri. Infl.New NCASI* 68360 Lbs/day
Pri Inf Turbidity 11 . 00 18 . 69 13 . 17
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day
Secondary Effluent 52633 Lbs/day 47737 39472 50032
#1 Sewer 1510 Lbs/day 2820
#2 Sewer 1665 PPM 847
2B Sewer 3900 Lbs/day 6965 4501 7837
3A Sewer 35470 Lbs/day 12838 12309 20228
#4 Sewer 492 PPM 342
Contaminated Condensate Sewer Values
Concentration 425. 00 mg/L
Flow 643. 13 GPM
Color 3279. 96 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 86. 00 mg/L
Flow 183. 98 GPM
Color 136. 88 Lbs/day 1186 1216 2251
5B Sewer 5990 Lbs/day 10508 7434 13581
CRP Color 4661.59 Lbs/day 4965
6A Sewer 4970 Lbs/day 16736 13671 23053
6A Sewer pH 3 .70
Pine Bleach 9540 Lbs/day 9627 10119 18011
19 . 14 Lbs/ton pulp 16 .32 17 .27 31 .44
Pine D1 310 Lbs/day 5539 6420 12538
Pine Eo 9170 Lbs/day 2687 2658 4254
Pine D2 60 Lbs/day 871
Hardwood Bleach 13600 Lbs/day 12753 12439 21190
16.50 Lbs/ton pulp 17 .25 14 .2 30 . 1
Hardwood Dl 6780 Lbs/day 5905 5292 13435
Hardwood Eo 6820 Lbs/day 6848 5875 11815
BFR Closure 53 .3 %
Unaccounted Color 21492 . 0 Lbs/day
Sewer Total 53997. 0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 13

Color Loadings to Mill Sewer Areas 30-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 101220 Lbs/day 65705 53400 81963
Pri.Infl.New NCASI* 98345 Lbs/day
Pri Inf Turbidity 18. 00 18.69 13 .54
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day
Secondary Effluent 51944 Lbs/day 47737 39472 50032
#1 Sewer 1280 Lbs/day 2820
#2 Sewer 266 PPM 847
2B Sewer 4300 Lbs/day 6965 4501 7837
3A Sewer 31900 Lbs/day 12838 12309 20228
#4 Sewer 501 PPM 342
Contaminated Condensate Sewer Values
Concentration 221 . 00 mg/L
Flow 556.57 GPM
Color 1476 . 02 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 49 .00 mg/L
Flow 227 .75 GPM
Color 235 .04 Lbs/day 1186 1216 2251
5B Sewer 8330 Lbs/day 10508 7434 13581
CRP Color 4494 . 04 Lbs/day 4965
6A Sewer 7180 Lbs/day 16736 13671 23053
6A Sewer pH 3 . 70
Pine Bleach 9560 Lbs/day 9627 10119 18011
23.35 Lbs/ton pulp 16.32 17 .27 31 .44
Pine D1 660 Lbs/day 5539 6420 12538
Pine Eo 8850 Lbs/day 2687 2658 4254
Pine D2 50 Lbs/day 871
Hardwood Bleach 15790 Lbs/day 12753 12439 21190
19 .40 Lbs/ton pulp 17 .25 14 .2 30. 1
Hardwood Dl 9170 Lbs/day 5905 5292 13435
Hardwood Eo 6620 Lbs/day 6848 5875 11815
BFR Closure 57 . 3 %
Unaccounted Color 47463 .0 Lbs/day
Sewer Total 53757 . 0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 14

Color Loadings to Mill Sewer Areas 31-Dec-02
Long Term In Control Upper Control
Sewer Area Daily Color Units Average Average Limit
---------- ----------- ----- ------- ------- -----
Primary Influent 37821 Lbs/day 65705 53400 81963
Pri.Infl.New NCASI* 36413 Lbs/day
Pri Inf Turbidity 33 . 00 18 . 69 14 .30
PI BMP LAL 70,322** Lbs/day
PI BMP UAL 78, 609** Lbs/day
Secondary Effluent 49691 Lbs/day 47737 39472 50032
#1 Sewer 400 Lbs/day 2820
#2 Sewer 265 PPM 847
2B Sewer 2460 Lbs/day 6965 4501 7837
3A Sewer 9450 Lbs/day 12838 12309 20228
#4 Sewer 214 PPM 342
Contaminated Condensate Sewer Values
Concentration 243 . 00 mg/L
Flow 796. 17 GPM
Color 2321 . 63 Lbs/day 3113 2435 4390
Combined Condensate Sewer Values
Concentration 11 . 00 mg/L
Flow 129 .61 GPM
Color 76.21 Lbs/day 1186 1216 2251
5B Sewer 10770 Lbs/day 10508 7434 13581
CRP Color 3463.21 Lbs/day 4965
6A Sewer 4860 Lbs/day 16736 13671 23053
6A Sewer pH 3 .50
Pine Bleach 8390 Lbs/day 9627 10119 18011
14 . 07 Lbs/ton pulp 16.32 17.27 31 .44
Pine D1 630 Lbs/day 5539 6420 12538
Pine Eo 7710 Lbs/day 2687 2658 4254
Pine D2 50 Lbs/day 871
Hardwood Bleach 14170 Lbs/day 12753 12439 21190
17. 19 Lbs/ton pulp 17.25 14 .2 30 . 1
Hardwood D1 8060 Lbs/day 5905 5292 13435
Hardwood Eo 6110 Lbs/day 6848 5875 11815
BFR Closure 62 .3 %
Unaccounted Color 9402 . 0 Lbs/day
Sewer Total 28419.0 Lbs/day
* New NCASI color test method which accounts for turbidity.
**The Primary Influent Lower Action Level (LAL) and Upper Action Level (UAL)
must be exceeded for two consecutive days before investigative or
corrective action is mandatory.
Figure 15

Overview - Mill Sewers Hill Filter Plant Erco Flows - Color- Fiber
Chemical Storage 0 10/14/200310:13:53 AM
0.32 MGD
F/L 2.47 MLbs Color Eo 4.4 MGD #19 PM
Flit ark Depoly 14.32 Tons Fiber
X- Rileyy leach 7.36 Mlbs Color D-1
Ch 9 Boiler Broke plant 1.19 MGD #20 PM
3.28 Tons Fiber
#2 Fiberline D
1
0. I 4.0 Delta T Tonss Fiber G
0.4 . _9GQ-Ac.3 E 17.6 River Temperature
�6a; S #1 F/L
.6 1Las Coi T
35 .L75 GaEo E Brown
F/L Swr —tT29 Tons Fiber R Stock #14 PM Area
S Washing
tr 0.72 WBL #12 PM
MGD
#11 RB 4 Evaps Power Power Boilers
Boiler 0.52
MGD
#10 RB
t=Ll)s..Capo #11 PM
Kilns 0.01 Tons Fiber
Syste s Off 1.22 Tons Fiber
1.44 MLbs Color,
3.77 MGD 22.91 Color Q TN
Re- 5.18
MG HW 2.49 O 8.89 D.O. @ F'vl
Ca at 5� Screen Tons Fiber
Room "27.
1.62 MGD 43230 #Color
54907 WTP 8
24.68 MGD Milt(Yesterday) 8 Turbidity #Color P409
120.22 MGD'RiverFlow' 1924MGD
#/Day TSS
Figure 16

i t> 1
Figure 17
+ I 1 1
COf'lDl.Cw 1'4I T'1 FLt�W 'EMP
Causticizing 689. 5. 19 77. ail:'
Recovery Boilers 3820. 1 , 55 102. ,r il.!! ;l)It
Evap Boil Out Sump 431 , � ; � �;•::r„
Evap Plorth Surnp 4. � �•i:Is ? ;.?,�
Evap South Sump 679. � ,,� ,
Digester 2798,
Hwd Wash 1267.
26 Sewer Dig/1FL 0, 1 , 02 179.
Area 2 545.
Stripper Feed 518.
Stripper Bypass 0. 00
�hlxt Scrn1
Stripper Sump 344 . -_---- --J
�Illi � IBIf� + �Vd 1Llii I ll�Il 11111kP , Prtli �ii �k; + llll �+l': ;i � .-nl,lil

ANOVA Factorial summary data
Based on daily data: 9/1/02 - 5/19/03
ANOVA results for Secondary Effluent Color >3.9 is <.05 is
(Only significant effects are listed) significant significant Targets
Main Effects F-Value P-value Used
BFR Closure 29.84 <.0001 75%
Pulp Mill Flow 11.447 0.0008 6 mgd
Contaminated Condensate Color 6.25 0.0131 2435 Ibs/day
Mill Flow 5.53 0.0196 24 MGD
Hwd Di Acid Use 4.58 0.0333 0.45 gpm
Pine Pre 02 PNs 4.33 0.0385 17
Interaction Effects
WL TSS* Pine Eo Color 11.998 0.0006
Closure Hwd PreBleach Conductivity'WL TSS ' Mill Flow 8.68 0.0036
Closure*WL TSS' Pine Eo 6.121 0.0141
PreBleach Cond. 'WL TSS * Mill Flow 5.77 0.0172
Hwd D1 Acid Use " Pine Pre 02 PNs 4.216 0.0411
Closure* Hwd PreBleach Conductivity 4.21 0.0414
Closure * Mill Flow 4.04 0.0457
ANOVA results for Unaccounted Color Targets
Main Effects F-Value P-value Used
Hwd PreBleach Conductivity 14.95 0.0001 600 umhos/cm
Contaminated Condensate Color 14.705 0.0002
Pulp Mill Flow 13.996 0.0002
WWTP Color Removal 9.889 0.0019 25%
Pine Eo Color 9.300 0.0025 2658 Ibs/day
Hwd D1 Acid Use 7.559 0.0064
Interaction Effects
WWTP Clr Rmvl * Hwd D1 Acid *Closure * Hwd PB Conductivity 13.707 0.0003
Contam. Condensate * Hwd PB Conductivity* Closure 11.189 0.001
Hwd D1 Acid *WL TSS * Hwd PB Conductivity 10.367 0.0015 _
Mill Flow* Closure 9.847 0.0019
WWTP Clr Rmvi*Closure * Hwd PB Conductivity 9.228 0.0027
Contam. Condensate* Hwd PB Conductivity 8.054 0.005
Pine Eo Color* Contam. Cond. Color* H PB Conduct. ' Closure 6.819 0.0096
Pine Pre 02 PNs *WL TSS 6.652 0.0106
Hwd PB Cond. * Closure 4.932 0.0274
Contam. Cond. Color* Closure 4.719 0.0309
Pine Eo Color* Contam. Cond. Color 4.395 0.0372
Pine Eo Color* WL TSS * Closure 4.247 0.0405
Pine Eo Color* Pine Pre PNs 'WL TSS 'Closure 4.145 0.043
For WL TSS, Target of 80 mg/I was used
Figure 18a

ANOVA Factorial summary data
Based on daily data:1/1/02 - 7/31/02
This Period represents GOOD SE Color Performance
ANOVA results for Secondary Effluent Color >3.9 is <.05 is
(Only significant effects are listed) significant significant
Main Effects F-Value P-value
Closure 16.555 <.0001
Interaction Effects
Pine Pre 02 PNs * Pulp Mill Flow 7.421 0.0071
Pine Eo Color* Closure * Hwd PB Conductivity 4.712 0.0313
ANOVA results for Unaccounted Color
Main Effects F-Value P-value
NONE
Interaction Effects
NONE
Figure 18b

Daily Color Monitoring Information
Report Date: Imin =
Primary Influent Color
Lbs/D
1w.wo'
wAw _______I______._ 1 I I I
_____ r___ ____r_______
I _
_ -A
_____L_______
I 1 I
�s.- _______I_ ____J_______J____ __ __ _ __
I I I I
_1
I I t I 1
x,wo
B9 LI] YA B9 IN Iryll 1N10
Secondary Effluent Color
Les/D
m,ow
I I I I I
I I I I I
I I I 1 I
I
I
]O.wJ _L. I _I J
I I I I I
S.M
bb Y1] YA 99 Ids le'II t0
i!
Unaccounted Color
Lbs/D
w..
I I I I I i
I I I 1 I
I I I I I
I I I I
I I I 1
tO.CN _L _
I I I I l
a
ze wa We vn Ia1 Ian m1e
NfrP Color Re1nMZmoval
%of Prima ent Color
w
I I I I I
I I I I
I I 1 I
I I I 1 I
I I I I 1
_I
I I I I
1 1 I I I
I I I I
a
se an cza vn ]1w Im1 Id1s
Secondary Effluent(Mll)Flaw
MGD
ss
I 1 I I 1
I 1 I 1 1
I I I 1 I
I I I I I
d8 YI] V30 gT Id< IWI1 IdIB
Figure 19

Daily Color Monitoring Information
Report Date: tN14rz003
Pulp Mill Flow
MGD
m
I 1 I I I
I I I 1
I I i I
1 I I
-i -1 I _f
I 1 I 1 I
x _ l _J _J _ _I _L
I I I I I
I 1 I I
0
% B'1] 4A Bbr lal ib 11 IaIB
DFR Closure
Percent
lro
I I I I I
I I I I
I 1 I
-y 1
I I 1 I I
I I I I
I I I I
I
I I I I I
I I I 1 I
B
% an am m lm rou Ime
Pine Pre-02 PN
xo
I I I I I
I 1 I I I
I I I I
I
_I _f _
I I 1
1 1
I I I I I
IB _1 _J _J I--------L
I I I I 1
I I I I I
% y,J YA q2! IW loll IbIB
Pine Post-02 PN
Ka aani and Lab
—PInB Post-D2 PN(Kelaarl�y-PNe Post-02 PN Qab)
le
v -L _J _I _L L _
1________I________I ____
-A 1
1] I
g ______ _____I_____ __L_______L_______
I 1 I I
p --------♦-------J--- -----I______ _ ______-F _ _______
4
B I I I I 1
B
% do em am lal ,an lale
Pine Eo Color
1000 LbsID
A
I I I I I
I I I I I
is _______ I I I _______I
I I I N--
B _______1_______ I__ I ____ L_ _____
I I
a
% al] 8'A 9Z! 1dB Itltt IaIB

Daily Color Monitoring Information
Report Date: lwla 000
Hardwood Pre-Bleach Conductivity
uMhas/cm
mm
I I I I
I
I I 1 I
,dL ------ -�---- I 1
I 1 ----I
1 1 1 I
I I 1 I
a
% al] %1 m lea Ian Iola
Hardwood DI Acid Use
GPM
I I I I
O.eO _r _1 -I -,
I I , I ,
6® -f _y _I _ I
I 1 I I
00 _l { _� _I
' I I I I
0A _L _1 _J _J _ _____
_I
I I 1 I I
ay
as tll] dS 92l IW Idll Wn,
While Dquor Insoluble Calcium-Ume lelns
m as CaO/L
sea
1 I 1 1 I
I 1 I I I
� T
I 1 I I
a
% an as as ro. mu Iola
Contaminated Condensate Color
Lbs/D
soon
1 I I 1
I 1 I 1 I
1 I I 1
I I I I 1
Iy] __.___ _a __ _ _-J- __ ____1_-___
1 I I I I
I 1 1 I 1
of
95 dl] YM q2! Itl. loll fd i]

Daily Color Monitoring Information
Report Date: 1W14=03
Hartlwood Quaternary Screen Rejects
GPM
m
I 1 I I I
1 I I I I
__ I
-� -T I_ -I
I I I 1 I
I 1
I I 1 I I
I I I 1
I I I I 1
0
4E tln � nv Ia. rou lase
Pine Quaternary Screen Rejects
GPM
xw
I 1 1 1 I
1 1 1 1 I
1 I I I I
I
IW ' ---- ' -I ' --I--
I 1 1
I 1 I I
I________I____ ____
I I I I I
1 I I I I
0
B9 B'1] gA yp fa. lal. Itl IB

Date: ########
Current Year: 2003
Current Month: 10 10/1/2003
Next Month: 11 11/1/2003
No. of Days-Current Month: 31
Target 1 Target 2
Target Daily SE Color(Lbs/D): 40'000. _42,000;
Total Monthly SE Color(Lbs): 1,240,000 11302,000
No.of Days SE Color SEColar No.of Days 'Balance' Updated 'Balance' Updated
Elapsed wp:seccl#.I MTD°Averag Remaining DallyLimit Daily, mill,
Date Lbs tbalD Lbs LbalD Lbs Lba10
10/1/2003 1 35,844 35, , 30 1,204,156 40,139 1,266,156 42=5
10/2/2003 2 42,528 39{186 29 1'.161,628 401056 1,223,628 42 194
10/3/2003 3 46,129 4.1,500 28 1,115,500, 39,839 1,177,500 421054
10/4/2003 4 70,086 4816.47. 27 1,045,414 3817,19 1,107,414 41d016
10/5/2003 5 46,495 48121B 26 998,919 r4201 1,060,919 0;805
10/6/2003 6 52,774 48,978 25 9461P145 3748446 1,008,145 4D,326
10/7/2003 7 42,611 48;067 24 903,533 3 77"ZA 965,533 401.231
10/8/2003 8 33,753 46;278 23 869,780 37j8jT7A 931,780 40;5j2
10/9/2003 9 29,977 44R66 22 839,803 38S1i7,3 901,803 40199,1
10/10/2003 10 37,049 431 6 21 802,754 B1226} 864,754 4;1,179
10/11/2003 11 48,433 44,153 20 754,321 816,321 40;816
10/12/2003 12 42,558 44;0'0 19 711,763 37,!461 773,763 40;•72,4
10/13/2003 13 43,230 43;959 18 668,533 - - 37�i1;41 730,533 40;585
10/14/2003
10/15/2003
10/16/2003
10/17/2003
10/18/2003
10/19/2003
10/20/2003
10/21/2003
1 0/2 212 0 0 3
10/23/2003
10/24/2003
10/25/2003
10/26/2003
10/27/2003
10/28/2003
10/29/2003
10/30/2003
10/31/2003

Section 3
Sewer Generated Color
Attached are two Duke University Graduate studies on sewer generated color
that were performed for the Canton Mill.
Historically, the effects of sewer generated color were seen on the acidic bleach
plant filtrates (D1 and D2 stages). It is now understood that the effects are more
far-reaching and largely influenced by a range of factors. Lab studies are being
organized to better quantify the color resulting from the interactions pinpointed in
the ANOVA studies.
The attached picture demonstrates the immediate increase in color when the
acidic sewer sample (6A) undergoes an increase in pH. The 6A sewer sample
had an initial pH of 3.7; it was increased to 10, then decreased to 7.8 to simulate
pH control at the Wastewater Treatment Plant. The color of the sample at a pH
of 7.8 was 682 true color units, compared to an initial color of 498. The same pH
adjustment was made on the 6A sewer sample the day before and the increase
on that day went from 402 true color units to 608 color units. The attached
studies show that sewer generated color can increase color up to 88%.
Information Submittal to EPA
Blue Ridge Paper Products Inc.
Canton, North Carolina
NPDES Permit#0000272
October 14, 2003

I
Sewer Simulation — pH adjusting 6A sewer sample to view
effect of sewer generated color
i
1
i
1
� I
I i
1
6A composite sewer sample with This sample was pH adjusted to 10,
Secondary Effluent sample NO pH adjustment. Initial pH was then back down to 7.8. The color of
3.7 and the color was 498 true color this sample was 682 true color units,
units. a 37% increase from the non-pH
adjusted sample.

1
A LABORATORY ANALYSIS OF THE COLOR REMOVAL MECHANISM ACROSS THE
WASTEWATER TREATMENT FACILITY OF A PULP AND PAPER MILL
CANTON,NORTH CAROLINA
by
Chad Michael Salisbury
Date:
Approved:
Dr. Gabriel Katul, Advisor
Dr.Norman L. Christenson,Dean
Master's Project in partial fulfillment of the requirements for the Master of Environmental Management
degree in the Nicholas School of the Environment of Duke University
1996

ABSTRACT: A twelve week study was conducted throughout the summer of 1995 at Champion
International Corporation's Canton,North Carolina mill and wastewater treatment plant in order to better
characterize the color removal mechanism occurring across the wastewater treatment plant Four distinct
areas of experimentation were conducted as part of this on-going research: adsorption isotherm, calcium
source, sewer generated color and bleach filtrate recycling (BFRT") simulation experiments. This report
presents the experimental methods and design for each area of experimentation as well as specific results,
discussions and conclusions. The findings are then presented in terms of their possible influence on the
future of the BFRTm technology for Champion as well as the pulp and paper industry as a whole. These
studies confirm that calcium is the significant factor in driving the color removal mechanism. At initial
color loads under 400 ppm, color removal efficiencies were modeled and found to be stable. Increasing
the temperature was found to increase percent color removal. The sewer generated color experiments
showed evidence that sewer generated and regular hardwood bleach color may be somewhat removable
across the wastewater treatment plant process. Finally, the BFRTM process will decrease color loads to the
wastewater treatment plant by recycling pine filtrates and color removal efficiencies should typically
remain above the projected 30%color removal level.

Table of Contents
1.0 Introduction I
I.1 Purpose 1
1.2 Background 2
1.3 Research Objectives 3
2.0 Experimental Methods and Design 5
2.1 General Methods 5
2.1.1 Biosolid collection and preparation 5
2.1.2 Color concentration measurement 6
2.1.3 Soluble Calcium concentration measurement 7
2.1.4 Total Suspended Solids measurement 8
2.2 Adsorption Isotherm Experimental Methods and Design 8
2.2.1 First Adsorption Isotherm Experiment- Week of 5/31/95 10
2.2.2 Second Adsorption Isotherm Experiment-Week of 615195 11
2.2.3 Third Adsorption Isotherm Experiment- Week of 6/12/95 11
2.2.4 Fourth Adsorption Isotherm Experiment-Weeks of 6/28/95, 8/1/95 12
2.2.5 EDTA Desorption of Color off the Biosolids 12
2.3 Calcium Source Experimental Design and Methods 13
2.3.l Original Calcium Source Experiment 13
2.3.2 Experiment Testing Effect of Temperature on Color Removal 16
2.4 Sewer Generated Color Experimental Design and Methods 16
2.5 BFRTm Simulation Experimental Design and Methods 18
3.0 Experimental Results and Discussion 21
3.1 Adsorption Isotherm Experiments 21
3.1.1 Results and Discussion 21
3.1.2 Conclusions 27
3.2 Calcium Source Experiments 28
3.2.1 Results and Discussion 28
3.2.2 Conclusions 32
3.3 Sewer Generated Color Experiment 33
3.3.1 Results and Discussion 33
3.3.2 Conclusions 35
iii

3.4 BFRTM Simulation Experiment 35
3.4.1 Results and Discussion 35
3.4.2 Conclusions 40
4.0 Research Conclusions: Implications of BFRTm 41
5.0 Suggestions for Further Study 43
6.0 Conclusion 45
7.0 Acknowledgements 45
APPENDICES 46
Appendix A 46
Appendix B 47
Appendix C 48
Appendix D 49
Appendix E 50
References 51
iv

1. Introduction
1.1 Purpose
During the summer of 1995, a set of laboratory experiments were conducted to better understand
the color removal mechanism occurring across the wastewater treatment facility of a bleached kraft pulp
and paper mill owned and operated by Champion International Corporation. These experiments are part
of an ongoing effort by Champion to gain a better understanding of the color removal processes at the
wastewater treatment plant. The importance of this work is its effort to predict the possible implications
of bleach filtrate recycling on color removal at the Canton mill.
This project will focus on the results from the four experiments conducted this summer. First, a
set of adsorption isotherm experiments were conducted to quantify the effect varying initial color loads,
calcium concentrations and biosolid concentrations have on color removal. The maximum capacity of the
biosolids to remove color was also analyzed. Second, the calcium source experiments determined the
significance of four sources of calcium (precipitated calcium carbonate (PCC), CaCIZ, Hardwood DI
filtrate and Lime mud) in the color removal process. The effects of temperature on the color removal
process were also tested for PCC and CaCIZ. The third set of experiments were the sewer generated color
experiments. These experiments set out to determine the maximum impact of sewer generated color on
color loads to the wastewater treatment plant and to see if sewer generated bleach color is removable
across the wastewater treatment plant. The final set of experiments were the BFRTM simulation
experiments to predict the possible impacts of recycling bleaching filtrates on calcium concentrations and
on color removal across the wastewater treatment plant.
This report supplies the reader with adequate background material to understand prior research
as well as the significance of this work. Research objectives are then clearly laid out. The methods
section provides a detailed description of the design for each experiment, analytical techniques used and
samples collected. The results and discussion section follows with graphs, figures and explanations of the
major findings of this work. This section is broken down by experimental area so the reader may easily
locate information of interest. Each experimental discussion is followed by a conclusion specific to the

area of research. Finally an overall research summary is provided in order to tie together all of the
findings directly pertinent to the BFRTM project. It is envisaged that this report will provide a valuable
resource not only to Champion International Corporation but to the pulp and paper industry as a whole.
1.2 Background
A bleached kraft pulp and paper mill owned and operated by Champion International
Corporation is located on the Pigeon River in the rural town of Canton,North Carolina. Color discharges
into the river have been of great concern to Champion, the states of North Carolina and Tennessee(which
is less than 35 miles downstream of the mill), and the people of the region. Although the Canton mill is
one of the lowest bleached kraft paper mill dischargers of color in the country, the low flow of the Pigeon
River combined with its mountain setting contribute to aesthetic questions on what color levels in the river
Champion should be required to meet. The current color standard is the equivalent of 50 standard color
units at the Tennessee/North Carolina border on a thirty day average basis (Champion, 1995). When
flows are low in the river,especially during the fall, achieving 50 standard color units at the border can be
difficult for Champion.
Champion has invested in technological modifications at the Canton mill as well as conducting
research on the color removal mechanism at the wastewater treatment plant in order to address this and
other environmental concerns. Beginning in 1989, Champion invested$330 million dollars in the Canton
Modemization Project to update mill facilities, of which $200 million went directly into environmental
improvements (Swann, 1995). Champion is currently in the process of demonstrating a technology
known as bleach filtrate recycling (BFRTM) on a full scale. BFRTm, developed and patented by Champion,
recycles bleach plant filtrates in-mill, thus decreasing the amount of dissolved organic material (including
color) sewered to the wastewater treatment plant from the bleach plant (Champion, 1995). The major
components of BFRTM are currently installed at the Canton mill. The major focus of this summer's
research was to predict what the effects of this technology will mean to color discharges to the Pigeon
River and to the mill's wastewater treatment plant color removal effect.
2

Several studies were conducted on color removal prior to the initiation of this research. The
results of these efforts were the building blocks for this current work. In the 1980's, Steve Stratton (1987)
of Champion suggested that calcium plays a significant role in color levels within the wastewater
treatment plant. This hypothesis was reinforced by research results by Amy McCord during the summer
of 1994. Most of this work is a continuation of the conclusions and suggestions made in her report.
McCord (1995) found that the color removal mechanism is a physical adsorption process which acts
primarily on color from unbleached sources, and that the presence of calcium is essential for this process
to occur.
1.3 Research Objectives
Based on the above technologies and research, the following project objectives were outlined and
studied:
1) to determine the effects varying soluble calcium levels,biosolids concentrations, and initial
color loads have on the color removal process;
2) to quantify the maximum capacity of color removal by the biosolids;
3) to determine how different calcium sources(precipitated calcium carbonate(PCC), CaCIZ,
Hardwood D 1 filtrate, and Lime mud)affect the color removal process under various calcium
dosages,pH, and temperature;
4) to determine if"sewer generated" color(i.e. color that is formed within the sewer system as
waste streams combine) is removable across the wastewater treatment plant;
5) to quantify the maximum impact of sewer generated color on color loads to the wastewater
treatment plant; and
6) to determine what impact the recycling of bleaching filtrates by BFRTm may have on calcium
concentrations and on color removal across the wastewater treatment plant. BFRTM includes a
process to remove certain metals, including calcium,from bleach plant effluent prior to
wastewater treatment.
The results obtained from studying each of these objectives may ultimately lead to predicting the
level BFRT" will improve color discharges to the Pigeon River. This research also clarifies many of the
questions which arose from McCord's research and poses new questions for future research efforts. This
research is intended to benefit the pulp and paper industry worldwide. Since few mills have to be
3

concerned with color to the degree that Canton is, most other mills are not performing such research. As
this nation moves forward in its protection of environmental quality, research such as this work can only
help the pulp and paper industry now and in the future.
4

2. Experimental Methods and Design
2.1 General Methods
Throughout this entire project, several analyses were conducted in each of the four areas of
experimentation. Samples were collected and tested at various times during the course of the experiments
for concentrations, in mg/1, of color, soluble calcium and total suspended solids. The methods for these
general analyses will be briefly discussed as well as the preparation of the biosolids used in each
experiment. The details of each set of experiments will be discussed in the next section.
2.1.1 Biosolid collection and preparation
A 3710 Portable ISCO Sampler by Isco, Inc. was placed at the outflow end of the wastewater
treatment plant aeration basin to be programmed to collect samples of mixed liquor suspended solids
(MISS). MLSS is a term referring to the combination of wastewater and biosolids contained in the
aeration basin (McCord, 1995). In preparation for each day's experiments a 24 hour composite sample of
MESS (400 ml collected every 20-30 minutes) was gathered the previous day. The 2.5 gallon composite
sample container was then allowed to settle for approximately one hour. Using an Erlenmeyer flask
connected to an air vacuum, as much of the liquid portion was siphoned away without disturbing the solid
material. This concentrated slurry of biosolids was then centrifuged to produce an ultra-concentrated
'cake' of biosolids. Sets of four 750 ml centrifuge bottles were centrifuged at 3000 rpm for 20 minutes in
an IEC Centra-8 Centrifuge by International Equipment Company, The number of sets centrifuged
depended on the number of reactors run each day. Typically three to four sets provided enough biosolid
composite for use. A glass jar containing all of the centrifuged 'cake'was refrigerated overnight. A total
solids test was performed on each batch of biosolid composite. An aluminum weighing dish was filled
with the composite, weighed and the wet weight recorded. The dish was then placed into a 107 degrees
Celsius oven overnight to evaporate all the water from the biosolid sample. The next morning the dried
composite was weighed and the dry weight recorded. Total percent solids was calculated by: [Wet weight
5

(g) -Dry weight(g)) /Wet weight(g) • 100. This value was used to calculate the amount of biosolids(in
wet grams) to be added to each reactor (assuming dissolved solid component of mass is negligible). The
equation is: (Biosolid mass added in wet grams) ' (percent total solids/100) =Target TSS (gfl) of reactor.
Since the aeration basin has an average mean cell residence time(MCAT) of eight days, the delay between
collection and use of the biosolid composite (typically 1-4 days) was deemed insignificant and the biosolid
composites used simulated aeration basin biosolids adequately.
2.1.2 Color concentration measurement
A standardized procedure was followed to measure the true color concentration of collected
samples. A detailed description of this procedure can be found in McCord (1995). Properly prepared
samples are compared to platinum-cobalt standard solutions with known color concentrations to achieve
color measurements. The procedure initially calls for the addition of a phosphate buffer to assist in the
adjustment of color samples to a pH of 7.6. However, as Carpenter (1986) points out, several industrial
studies have indicated that the phosphate buffer could possibly interact with calcium contained in the
samples and interfere with the test results. Therefore the buffer was not added to any samples obtained
from reactors which were dosed with calcium. The buffer was added to undosed samples. Once the color
samples were filtered through a 0.8 micron porosity membrane filter by Millipore to remove turbidity, the
samples were measured for transmittance using a spectrophotometer or refrigerated overnight for
measurement the next morning.
Two different spectrophotometers, a Spectronic 20 and 21, were used depending on the intensity
of color in the samples. The Spectronic 20 is standardized using Pt.-Co. standards of 100, 500 and 1000
standard color units (scu) while the Spectronic 21 is standardized with 10, 100 and 500 scu standard
samples. The Spectronic 21 is used for low color range samples and gives more precise color readings.
Once the spectrophotometer has been standardized, the color samples were tested for light transmittance.
True color measurements in scu (or mg/l Pt.-Co. units) are obtained using a conversion from
transmittance to color concentration(NCASI, 1971).
6

2.1.3 Soluble Calcium concentration measurement
The second important variable for which samples were collected and tested was soluble calcium
concentration in mg/l. The work of Stratton (1987) and McCord (1995) prompted this analysis. A
Mettler DL70 ES titrator was utilized to determine the soluble calcium concentrations of each sample.
The titrator has a program to determine the hardness (mg/I of CaCO3) of a solution by titrating it with
EDTA(tetra sodium ethylene diamine tetra acidic acid). The principles of the titration procedure are the
same as those found in Section 2340 B. of Standard Methods for Water and Wastewater (1992). The only
difference is that an electronic phototrode in the titrator replaces the human eye in determining the change
in solution from wine red to blue, indicating that all the calcium in the sample has been complexed
(Pryately, 1995). It is important to note that hardness consists of both calcium and magnesium ions, but
magnesium is assumed to be insignificant because it is present in such small concentrations in the samples
collected at the wastewater treatment plant Therefore all of the hardness measured from the titrator is
assumed to be due to calcium ions. The computer within the titrator then calculates the concentration of
CaCO3 (mg/1)from the volume of EDTA and sample titrated and a printout is generated.
The titrator must first be calibrated according to set procedures for the hardness titration program
(Mettler manual, 1993). Although the details of the calibration procedures are not important here, the
EDTA and Hardness standards calibration steps must be done within +/- 5% or else be redone until so.
This ensures accuracy within standard error for the results. Once the machine is calibrated the titrator is
then programmed for the hardness test. The calcium samples were obtained by siphoning off 20-30 ml of
decant from each experimental reactor at various times throughout each experiment. It was important to
remove as much of the biosolids from the decant so as not to shift the established equilibrium between the
biosolids, calcium ions and color bodies within the samples. Often filtration through glass fiber filter
paper was used to remove the biosolids, especially when samples were gathered prior to completion of the
experiment. Samples not tested the day they were collected were refrigerated until the titrator was
7

operated. Five milliliters of calcium sample was placed in each titrator cup and duplicates nin to check
the reproducibility of the machine. Soluble calcium concentrations in mg/I of Ca' was calculated from
hardness (based on the assumption of little magnesium) using basic mole balance principles. Forty
percent of the mass of CaCO3 is attributed to the calcium ion. Therefore hardness values were converted
to soluble calcium concentrations by a factor of 0.40.
2.1.4 Total Suspended Solids measurements
The final general laboratory analysis conducted on samples was for total suspended solids (TSS)
in mg/l. The procedure for this analysis is Section 2540 D. of Standard Methods (1992). Most samples
were placed in a 107 degrees Celsius oven for only one hour, which in wastewater treatment samples is
standard procedure to remove all the water without degrading the organic content of the samples. TSS
samples run at the end of the day were kept overnight in the oven and weighed the next morning. TSS
was calculated by subtracting the dried weight (g) from the wet weight (g) and dividing by the volume
filtered. TSS concentrations were most important in generating the adsorption isotherm curves whereas
these values provided a check of the reactor target TSS concentrations in the other areas of
experimentation.
2.2 Adsorption Isotherm Experimental Methods and Design
A series of bench-scale experiments were conducted throughout the twelve week research period
based on the results in McCord (1995) and Koelsch (1994). Conclusions from their work showed that
calcium, biosolids and color level are all statistically significant in explaining the color removal
achievable for black liquor color. It is this result that motivated the four adsorption isotherm experiments
conducted throughout the summer. The following two questions were the focus of the analysis:
1) What happens to color removal efficiency as:
i) biosolids concentrations vary about the wastewater treatment plant aeration basin
average of 2500 ppm;
8

ii) calcium dosages to reactors change;and
iii) initial color concentrations in reactors change?
2) What is the maximum capacity of the biosolids to remove color?
Positive adsorption in a solid-liquid system results in the removal of solutes from solution and
their concentration at the surface of the solid, to such time as the concentration of the solute remaining in
solution is in a dynamic equilibrium with that at the surface (Weber, 1972). The solute dealt with in this
work is color bodies and the solid surface is the biosolid (bacteria) cell walls. As mentioned above
biosolid, calcium and initial color concentrations all shift this equilibrium, the direction depending on the
concentration of each with respect to the levels of the others. The preferred form for expressing this
equilibrium is the adsorption isotherm which is a plot of the quantity of solute adsorbed per unit weight of
solid adsorbent versus the concentration of solute remaining in solution (Weber, 1972). The adsorption
isotherms created in this report are plots of color adsorbed (mg) per gram of TSS versus measured final
color (mg/l platinum cobalt units). Adsorption isotherms generated by Koelsch (1994) depicted the
desorption portion of the curve only. Thus a goal of this work was to generate data for the entire curve in
order to better quantify the color removal mechanism occurring across the wastewater treatment plant.
Four separate adsorption isotherm experiments were run throughout the summer with each
additional experimental design being built upon the previous experiments. The fourth adsorption
isotherm experiment is the most significant of the four. The majority of the important results and
conclusions arise from the fourth experiment. However, experiments one through three provide
interesting insight to the color removal process as well as quantifiable support for this area of research.
All four experiments will be discussed and their relevance explored.
All four experiments were conducted using two jar stirring assemblies with magnetic stirrers.
Each reactor contained 1000 ml of black liquor solution prepared to target initial colors. Experiment
specific concentrations will be discussed in the following methodology sections. A sample of Black
Liquor(BL) stock solution (half from the Pine digestors and half from Hardwood digestors) was gathered
on the Monday of each week in which the experiment was conducted and its color measured and recorded.
Note the Black Liquor stock is extremely high in color and had to be diluted 1:4000 to be measurable on
9

the spectrophotometer. A five gallon batch of color solution to be used in the reactors was prepared on
Monday by combining predetermined volumes of deionized water and 10:2000 diluted Black Liquor stock
according to the mass balance formula:Volume(BL)+Concentmtion(BL)=Volume(Mix)•Concentration
(Mix). The volume of deionized water added equaled Volume (Mix) - Volume (BL). Once prepared and
well mixed, the five gallon batch was pH adjusted to between seven and eight to mimic the conditions in
the wastewater treatment plant. The color of the batch was then measured and recorded.
A predetermined dosage of calcium was added to each reactor, based on mass balance
calculations, as either 1000 ppm CaNO3 or 5000 ppm CaC12. The biosolids were added to each reactor in
wet grams. Target TSS concentrations were based on percent total solids calculations mentioned above.
Once all three reactor ingredients were added, the reactors were mixed for 6.5 hours, which is the mean
residence time for the wastewater treatment plant aeration basin (Datastream, 1995). After the
experiments were completed the reactors were permitted to settle for about an hour, after which samples
were gathered for color(100 nil), TSS(20ml), and soluble calcium (20-30 nil) and as described above.
2.2.1 First Adsorption Isotherm Experiment - Week of 5/31/95
The first adsorption isotherm experiment consisted of eight reactors. Reactors one through seven
contained 30.1, 41.2, 50.8, 61.28, 70.4. 79.7 and 100 wet grams of biosolids respectively. The biosolid
composite used was 7.96% total solids by weight. Reactor eight served as a control jar in which no
biosolids were added. The initial color of the WBL solution was 449 scu. After 15 minutes of mixing,
forty milliliters of 1000 ppm CaNO3 were added to reactors four through seven only, resulting in a reactor
calcium concentration of 40 mg/l, and the time was recorded as time zero. Due to the large amount of
biosolids added to reactors 4-7 a portion of the mixture was removed to reduce the volume to 1000 nil.
After one hour 200 ml samples were taken from each reactor. After 2.5 hours from time zero, 200 nil
samples were again taken from each reactor and the experiment stopped. Each sample was tested for
color, calcium and TSS.
10

2.2.2 Second Adsorption Isotherm Experiment- Week of 6/5/95
The second adsorption isotherm experiment refined the design of the first experiment by
subjecting the biosolid/WBL solution to both 40 and 60 mg/l concentrations of soluble calcium added as
1000 ppm CaNO3. Eighteen reactors were run with nine being run on each of two days. Reactors 1-7
and 8-14 contained 21, 30, 40, 50, 60, 70, 85, and 100 wet grams of biosolids respectively. All four
controls (C-1 through C4) contained no biosolids. Ten gallons of WBL solution were prepared with an
initial color of 431 scu. Reactors 1-7 and C-1 were dosed with 1000 ppm CaNO3 to produce a calcium
concentration of 40 mg/I. Reactors 8-14 and C-3 were dosed for a reactor calcium concentration of 60
mg/l. C-2 and C-4 were not dosed with calcium. These two controls served to check the color consistency
of the WBL solution while the other controls served to isolate the effect calcium additions have on the
WBL solution, if any. The reactors were mixed for 6.5 hours using the same apparatus as in the first
experiment. After the addition of calcium, samples were gathered for color and calcium concentrations
prior to calcium addition, after one hour of mixing and after 6.5 hours of mixing.
2.2.3 Third Adsorption Isotherm experiment- Week of 6/12/95
The third adsorption isotherm experiment was conducted in the same manner as the second
experiment with the following exceptions:
1) Four sets of four reactors with 20, 30,40 and 55 wet grams of biosolids added; and
2) Reactors 14 (set 1)not dosed with calcium, 5-8 (set 2)dosed with 40 mg/l of calcium added
as 5000 ppm CaCIZ, 9-12 (set 3)dosed with 60 mg/l and 13-16(set 4) dosed with 80 mg/l.
Set 1, set 2 and a control (C-1)were run for 6.5 hours on 6/12/95 and set 3,set 4 and C-2 were run for 6.5
hours on 6/13/95. The initial color of the WBL prepared solution was 414 scu. Samples for color,
calcium, and TSS were gathered only after the 6.5 hours of contact time was completed. The wet grams
of biosolids were reduced from the second experiment because the higher concentrations of biosolids were
11

well above the range of TSS at the wastewater treatment plant. Higher dosages of calcium were added in
order to quantify the effect of increasing calcium levels in the flow into the wastewater treatment plant
aeration basin.
2.2.4 Fourth Adsorption Isotherm Experiment -Weeks of 6128/95, 8/1/95
The final adsorption isotherm experiment was developed with the intentions of creating a
complete adsorption isotherm with both the adsorptive and desorptive limbs of a traditional isotherm. The
first three experiments were unable to achieve this goal while the fourth achieved it. The key factor not
present in the previous experiments was to vary the initial color concentration of each reactor. The
biosolid concentration was held constant at 2500 ppm to target the average TSS concentration at the
wastewater treatment plant. Biosolids were added to each reactor plus the controls. The wet grams of
biosolids added to achieve this target were based on the percent total solids value for each biosolid
composite. The procedure for this calculation is described above under general methods. Using mass
balance calculations described above, the appropriate dilutions of the WBL solution were calculated for
each reactor. Initial color concentrations varied between 100400 scu with the exception that the 6/28/95
reactors varied up to 565 scu. Seven reactors and a control (no calcium added)were run on each of the
three days that the experiments were conducted. The reactors on 6/28195 were dosed with 40 mg/I of
calcium added as 5000 ppm CaCh, 8/1/95 dosed with 60 mg/I and 8/2/95 dosed with 80 mgll. After 6.5
hours of mixing the reactors were sampled and tested for color,calcium and TSS concentrations.
2.2.5 EDTA Desorption of Color off the Biosolids
The four adsorption isotherm experiments provide a measure of the capacity of the wastewater
treatment plant biosolids to adsorb color from solution. These biosolids, in addition, already have a
certain amount of color attached to their cell walls which strongly influences how much more color can be
adsorbed. Thus to quantify the total capacity of the biosolids to adsorb color from solution, the amount of
color already adsorbed needs to be determined. Following desorption procedures developed by McCord
12

(1995), EDTA was used to desorb color off the biosolids used in the fourth adsorption isotherm
experiment. The recipe used was to combine 900 ml of deionized water, 100 ml of O.01M EDTA and
2500 ppm of biosolids to a reactor and mix for 6.5 hours. After 6.5 hours, the color of the solution was
measured and the desorbed color calculated in milligrams of color desorbed per gram of TSS. This
number was then added to the peak adsorption of the generated isotherm to determine the maximum
capacity of the biosolid sample to adsorb color.
2.3 Calcium Source Experimental Design and Methods
The second area of experimentation conducted over the summer of 1995 was the calcium source
experiments. The basic premise of this experiment was to test the effect the form of calcium has on the
color removal process (i.e. low soluble forms of calcium vs. soluble forms of calcium). Four sources of
calcium were tested for their adsorptive effects on the color removal process. The four sources were:
1) Precipitated Calcium Carbonate(PCC)
2) Soluble calcium added as 5000 ppm CaC12
3) Hardwood D I filtrate
4) Lime mud
PCC and Lime mud are both solid, low solubility sources of calcium while CaC12 and Hardwood D1
filtrate are both highly soluble sources. An additional experiment was conducted to test the effect of
temperature on the color removal efficiency of a solid, low solubility calcium source (PCC) and a soluble
source of calcium (CaCl2). The experimental design is the same, however it was conducted
simultaneously at room temperature and at 35 degrees Celsius in a waterbath. The following
experimental design was performed on each source of calcium listed above.
2.3.1 Original Calcium Source Experiment
The experimental design was developed by Martin Steinberg of Champion's Corporate
Technology department in West Nyack,New York(Steinberg, 1995). The design is a two-factorial design
13

with a centerpoint with the variables being pH(levels=7, 7.5 and g)and calcium dosage (levels=25, 40
and 55 mg/1) as seen in Figure one. The centerpoint is run in duplicate while two controls are run each
day using CaC12 as the calcium source and the centerpoint conditions in order to measure day and source
effects. Additional controls were run in which the reactors contained only WBL solution and biosolids to
estimate the "reserve capacity" of the biosolids.
Figure 1: Two-Factorial Experimental Design
for Calcium Source Experiment
g .................................._................................................................................................................................................
8.5
8
pH 7.5 •
7 • •
6.5
6
20 25 30 35 40 45 50 55 60
Calcium Dosage mg/I
Each day a different source of calcium was run under design conditions. The six design reactors
and two CaC12 control reactors consisted of 1000 nil of WBL solution (color approximately 325 scu) and
2500 ppm TSS added as wet grams of biosolids and were mixed for 6.5 hours using ajar stirrer apparatus.
The pH of each reactor was adjusted using hydrochloric acid (HCI) and/or sodium hydroxide (NaOH)
prior to addition of the calcium. Samples for color, calcium and TSS concentration were gathered after
6.5 hours of contact time.
The PCC sample, collected on 6115195, was 19.4% as CaCO3 (194,000 ppm CaCO3). Using
mole balance calculations, the concentration of W2 was 77,600 ppm. The volume of PCC slurry added to
14

each reactor was calculated using mass balance calculations with the final mixture volume equal to 1000
ml, calcium concentration in PCC equal to 77,600 ppm and the mixture calcium concentrations from the
reactor design. The PCC experiment was run on 6/20/95. The CaC12 trial used a 5000 ppm solution of
CaCl2 (prepared 8/29/94) as the source of calcium and was run on 6/21195. Volumes of CaCIZ were
calculated using mass balance calculations as in the adsorption isotherms work.
The Hardwood DI trial was run on 6/22195 and was the most complicated to set up. The
difficulty arises from the fact that the Hardwood DI calcium source is also a source of color that must be
considered in preparing the reactor solutions. The color of the Hardwood DI was measured as 450 scu
while Champion's West Nyack Corporate laboratory determined the calcium concentration to be 137
mg/l. The composite Hardwood D1 sample was created by mixing samples collected over the nights of
6/16/95, 6/17/95 and 6/18/95. The following two mass balance equations were solved simultaneously to
arrive at the volumes of WBL, Hardwood D1 and deionized water to achieve a target color of 325 scu and
target calcium dosage level:
1) Volume Hwd DI * Calcium Hwd D1 = 1000 ml *Target Calcium level
2) Volume WBL* Color WBL=(1000 ml * 325 scu)-(Volume Hwd DI*Color Hwd D1)
The volume of Hardwood D1 (ml) is first calculated and this value is entered into Equation 2 above to
calculate the volume of WBL solution (ml). The sum of these values is subtracted from 1000 ml to get the
volume of deionized water(ml). This produces a recipe for 1000 ml batches of reactor solutions for each
design point. Table one illustrates the calculations for the blends of each reactor point. A double batch of
TABLE I: Reactor Blends for Hardwood DI Calcium Source Experiment
Blend Hwd D1 WBL Hwd D1 Volume Volume Volume Blend Blend
Calcium Calcium Color Color Hwd D1 WBL DI water target color Actual color
mg/I mg/I scu scu ml ml ml scu scu
25 137 1342 450 183 181 636 325 370
40 137 1342 450 292 144 564 325 384
55 137 1342 450 401 108 491 325 315
15

each solution was mixed so as to have enough for the entire experiment. Each reactor was then pH
adjusted to between seven and eight pH units. After 6.5 hours of mixing samples for color, calcium and
TSS were collected and analyzed.
The final calcium source tested was lime mud which was 66% solids (100% of solids CaCO3).
Through mole balance calculations the concentration of calcium was 264 mg calcium per gram of lime
mud. The lime mud was collected on 6/14/95 while the source experiment was conducted on 6/23/95.
The mass of lime mud (g) added to each 1000 nil reactor was calculated by dividing the desired calcium
concentration by 264. The remaining process is the same as for the other source trials.
2.3.2 Experiment Testing Effect of Temperature on Color Removal
The above experimental design was repeated simultaneously at room temperature (26.5 degrees
Celsius) and in a 35 degrees Celsius waterbath. The waterbath was included in this experiment in order
to simulate wastewater treatment plant aeration basin conditions and determine the effect that temperature
has on color removal efficiency. PCC and CaClz were chosen for this experiment to represent both a
soluble and low solubility form of calcium. The room temperature reactors were set up the same as above.
The waterbaths were first equilibrated at 35 degrees Celsius before the experiment began. After the
reactor contents were properly prepared, they were placed in the waterbaths and aerated using aeration
stones. Both temperature experiments were mixed for 6.5 hours and appropriate analyses conducted. The
same calcium stocks were used as in the original experiments. The PCC trial was performed on 6/29/95
and the CaCI=on 7/11/95.
2.4 Sewer Generated Color Experimental Design and Methods
The third set of experiments conducted dealt with two questions concerning the phenomenon of
sewer generated color (SCG). SCG is the phenomenon observed when the measured color of the
combined waste flows is greater than that which can be accounted for by the color contributions of all the
individual waste streams (Caron, January 1992). This phenomenon has been explained by chemical
16

reactions that occur as the acid sewer(Pine and Hardwood D 1 filtrates) is introduced to the sewer system.
These acid sewers react under high pH (10-11) and high temperature (120 °F) within the sewer system
(Caron, February 1992). SGC accounts for approximately 20% of the estimated color at the sum of the
sewers (Koelsch, 1995). The two questions raised by[his phenomenon are:
1) What is the maximum impact of sewer generated color on color loads to the
wastewater treatment plant?
2) Is sewer generated bleach color removable across the wastewater treatment plant?
The experimental design for this research consisted of two stages: (1) raising the pH of Pine and
Hardwood DI filtrates to pH = 11 and exposing them to 120 degrees Fahrenheit temperatures for 30
minutes to generate new color and (2) subjecting sewer generated and regular pine and hardwood bleach
filtrates to wastewater treatment plant conditions(Koelsch and Salisbury, 1995).
Stage one was designed to simulate sewer generated color conditions present in the sewers prior
to entering the wastewater treatment plant. The experiment was conducted on 7/12/95 and repeated again
on 7/13/95. On the morning of each day fresh composites of Pine D1, Pine Eo, Hardwood DI and
Hardwood Bo filtrates, sampled throughout the night before, were gathered. Initial color and calcium
samples were collected and tested for each filtrate. A portion of the Pine and Hardwood DI filtrates was
then subjected to sewer generated color conditions. These "SGC" filtrates were increased to pH = 11
using NaOH and placed in an equilibrated 120 °F waterbath for 30 minutes to simulate the SGC
conditions. Next, these SGC Pine and Hardwood Dl filtrates were then sampled and tested for color and
calcium concentrations. At this stage of the experiment, the maximum impact of sewer generated color
on color loads to the wastewater treatment plant was quantified.
Stage two of the experiment was to create sewer generated and regular bleach, pine and
hardwood filtrate blends and subject them to wastewater treatment plant conditions. The filtrate flows
were blended volume proportionally based on comprehensive mill data in Datastream (1995). Table two
illustrates the blend ratios used. For each blend samples were collected prior to wastewater treatment
plant conditions for color and calcium concentrations. Each reactor contained 750 ml of a filtrate blend
serving as the color source, 40 mg/I dosage of calcium added as 5000 ppm CaC12 and 2500 ppm of
17

biosolids. Prior to adding the calcium and biosolids, each reactor was pH adjusted to pH=7.5 to simulate
conditions in the wastewater treatment plant aeration basin. The reactors were mixed by aeration in a 35
degrees Celsius waterbath for 6.5 hours. Upon completion of the each trial, samples were collected and
tested for color and calcium concentrations. By comparing the color removed from sewer generated and
regular filtrate blends,we were able to quantify how much sewer generated color was removed.
TABLE 2: Filtrate Blend Ratios for the Sewer Generated Color Experiment
Blend Filtrate %of Total 750 ml Blend
Bleach Filtrate flows Pine D1 54.3% 407
Pine Eo 2.5% 18
Hwd D1 26.6% 199
Hwd Eo 16.7% 126
Pine Filtrates only PineDl 95.7% 717
Pine Eo 4.3% 33
Hwd Filtrates only Hwd D1 61.3% 460
Hwd Eo 38.7% 290
2.5 BFRTM Simulation Experimental Design and Methods
The final set of experiments were the BFRTM simulation experiments designed to predict the
impact that recycling bleaching filtrates may have on calcium concentrations and color removal across the
wastewater treatment plant BFRTM, developed and patented by Champion, recycles bleach plant filtrates
in-mill thus decreasing the amount of dissolved organic material (including color) sewered to the
wastewater treatment plant from the bleach plant (Champion, 1995). The major components of BFRTTM
are currently installed at the Canton mill. Maples et. al. (1994) provide a detailed description of the
BFRTM process. The concern of this report is on predicting the possible effects of this technology on color
discharges to the Pigeon River.
The experimental design for the BFRTM experiments was initially created during a Canton color
removal project planning meeting(Henderson et. al., 1995) in February of 1995 and later refined during a
July 13, 1995 color teleconference (Hilleke et. al., 1995). Table 3 shows the experimental design for each
18

two day trial. The variables selected for this design are calcium level, color level, blend ratio (ratio of
brown color to hardwood filtrate color) and biosolids concentration. The centerpoint reactors represent
the expected mean values for each variable under BFRTM conditions while the other values represent
expected ranges above and below the mean based on calculations by the BFR774 Development Team at the
Canton mill. The biosolids concentrations used bracket averages from the wastewater treatment plant.
The experiment was conducted over a two week period whereby two trials were done each week. Each
TABLE 3: Experimental Design for the BFRTM Simulation Experiment
Jar# Day # Block Calcium Level mg/l Color Level Blend Ratio Biosolids conc.
(Added CaC12) mg/l (%Brown Color) mg/l
1 1 1 45 380 85 2500
2 1 1 45 380 85 2000
3 1 1 45 380 55 2500
4 1 1 45 380 55 2000
5 1 2 45 200 85 2500
6 1 2 45 200 85 2000
7 1 2 45 200 55 2500
8 1 2 45 200 55 2000
9 1 Center pt 35 290 70 2250
10 1 Center pt 35 290 70 2250
1 2 3 25 380 85 2500
2 2 3 25 380 85 2000
3 2 3 25 380 55 2500
4 2 3 25 380 55 2000
5 2 4 25 200 85 2500
6 2 4 25 200 85 2000
7 2 4 25 200 55 2500
8 2 4 25 200 55 2000
9 2 Center pt 35 290 70 2250
10 2 Center pt 35 290 70 2250
trial required two'consecutive days for completion. Trials 1 and 3 were performed using biosolids
whereby color was first desorbed prior to centrifugation. The same general preparation procedure was
followed with the exception that the decanted bicsolid composite was filled to the top of the collection
container with distilled water and aerated for two hours. The "desorbed" biosolids were then permitted to
19

settle and the procedure continued. Trials 2 and 4 used current wastewater treatment plant biosolids
prepared using the original general procedure. The desorption trials were perforated in order to
equilibrate the biosolids to lower color loads expected after BFRTM has been operational for awhile. The
non-desorption trials simulate the conditions existing now in the mill while BFRTM is starting up.
On the Monday of each week new Hardwood DI and Eo samples and black liquor sample were
collected and tested for color as previously described. A ten liter batch of hardwood filtrate (61.3%DI
and 38.7%Eo) was prepared using the collected samples. Next, ten liter batches each of 85%, 70% and
55% brown color blends of brown and hardwood bleach color were prepared using mass balance
calculations targeting for an initial color of 500 scu. Table 4 lists the volumes used to
TABLE 4: Sample Blend Ratio Table for
BFR Simulation Experimental Reactors (Week Two Blends)
Blend Hardwood WBL color Blend Volume Volume Volume Blend
Ratio Filtrate (10:2000) final color Hardwood WBL DI water Final
%brown Color Dilution (theoretical) filtrate dilution added color
color scu scu scu nil nil nil scu
85% 972 1335 500 772 3184 6044 562
70% 972 1335 500 1543 2622 5835 499
55% 972 1335 500 2315 2060 5625 562
create each blend. The color level in each reactor was achieved using mass balance calculations for 750
nil final volumes and the target color level listed in table 3. Biosolids were added to each reactor in wet
grams according to general procedures and dosed with 5000 ppm CaC12 as listed in Table 3. The reactors
were then placed in a 35 degrees Celsius waterbath and aerated for 6.5 hours after which samples were
collected and tested for TSS, color, soluble calcium and total organic carbon (TOC) concentrations. The
TOC analyses (see Appendix E) were conducted by the BFRTm Development Team. Initial and final
colors were then measured and recorded for each reactor.
20

3. EXPERIMENTAL RESULTS AND DISCUSSION
3.1 Adsorption Isotherm Experiments
3.1.1 Results and Discussion
The results of the first three adsorption isotherm experiments overlap and lend detailed support
to the results of the more thoroughly designed fourth experiment. The first three experiments build upon
one another and allow for a detailed analysis of the effects of varying biosolid and calcium concentrations
on the color removal process. The results in these experiments make interpretation of the fourth, and
most significant adsorption isotherm experiment, much more organized.
Although the first adsorption isotherm experiment was a trial experiment performed to determine
the design criteria of the reactors used in the other isotherm experiments, one particular result stands out.
Figure 2 is a graph of measured color(mg/l Pt.-Co. units)vs. TSS (g/1) after one and 2.5 hours of mixing.
FIGURE 2: Graph of Measured Color (mg/1 Pt.-Co. units) vs. TSS (g/1)
for the First Adsorption Isotherm Experiment
Adsorption Isotherm 5131195-Measured Color After 1 and 2.5 hre of Contact
(Initial BLColor was 499 scu's)
700
600 .............................. ........ ......................................_.....................................................................
m 500.� .. No Soluble G edded ................. ........................................................................................................
0
0
u 400
n
E 300
..................................................................... ............................................................................................
c 200 40 mg/I Soluble Ca added
u ..........................................................................
100 ............................................................................... ......
0
1.94 2.27 2.00 3.31 4.O0 5.46 7.36
TSS 9n
t Final Color 1hr Final Color 2.5 hr
21

Note that in reactors without a calcium dosage, color is desorbed off the biosolids while in the reactors
dosed with 40 mg/1 of calcium color a large amount of color is adsorbed by the biosolids(see Table 5).
TABLE 5; Adsorption and Calcium Results from First Adsorption Isotherm Experiment
Hiosolids Calcium Color removed Color removed Soluble Ca Soluble Ca
Reactor TSS (g/1) Dose(mg/1) %after I hr. %after 2.5 hr mg/I after 1 hr mg/1 after 2.5 hr
1 1.94 0 -26% -16% 10.49 15.71
2 2.27 0 -24% -14% 11.66 15.46
3 2.60 0 -20% -14% 11.28 16.14
4 3.31 40 90% 78% 0 0
5 4.06 40 86% 81% 0 0
6 5.46 40 81% 79% 0 0
7 7.36 40 81% 79% 0 0
--Control 0 0 0% 0.928 1.76
4-duplicate 3.31 40 84%
There is a slight decrease in color adsorbed as TSS increases in the dosed reactors. The most interesting
data collected in Table 5 are the soluble calcium concentrations measured after one and 2.5 hours. In the
dosed reactors, soluble calcium concentrations after one and 2.5 hours of mixing are zero thus lending
extreme support that calcium concentrations in solution are the driving mechanism for the color removal
process being analyzed. This result was unable to be recreated in the other experiments, but still should
not be discounted.
The results of the second and third adsorption isotherm experiments can be reviewed together
due to their similar designs. The second isotherm experiment repeats the first with the exception of
dosing all reactors with calcium levels of both 40 and 60 mg/l of calcium. The third isotherm experiment
repeats the second with the exceptions of eliminating the four highest TSS reactors and dosing all the
reactors with 40, 60 and 80 mg/1 of calcium. Table 6 lists the results for the second adsorption isotherm
experiment while Table 7 lists the results for the third. To interpret this data one must remember the
concept of the biosolids having a "color history". This history is the fact that the biosolids used have
already been exposed to color bodies within the aeration basin. The control reactors in which the
biosolids were not dosed with calcium allow for a quantification of the color removal reserve capacity on
the biosolids. The biosolids used in the third experiment all desorbed color when exposed to
22

TABLE 6: Adsorption and Calcium Results from 2nd Adsorption Isotherm Experiment
Color Adsorbed
Ca dose Color(scu) %color removal Soluble Ca(mg/1) TSS per gram TSS
Reactor (P m) t� t=1 t--6.5 after 6.5 hrs t� t=1 t=6.5 (9/1) t=1 t=6.5
1 40 464 359 306 29% 16 59 58 1.56 46.0 79.9
2 40 456 321 235 45% 20 61 59 2.18 50.5 89.9
3 40 447 314 175 59% 19 54 53 2.59 45.2 98.8
4 40 414 242 143 67% 15 59 51 3.47 54.5 83.0
5 40 336 285 175 59% 14 46 45 4.46 32.7 57.4
6 40 314 256 143 67% 17 56 48 5.34 32.8 53.9
7 40 398 263 156 64% 19 56 46 5.82 28.9 47.3
8 60 464 169 61% 10 84 1.43 183.2
9 60 456 156 64% 11 84 1.66 165.7
10 60 447 162 62% 11 82 2.05 131.2
11 60 431 130 70% 14 78 3.38 89.1
12 60 423 I30 70% 16 78 4.16 72.3
13 60 398 137 68% 16 74 5.08 57.9
14 60 383 143 67% 18 70 5.96 48.3
71 0 431 414 4 1 4% 0 39 38 0.01 2833.3 2833.3
C-2 0 431 431 431 0% 0 0 0 0.01 0.0 0.0
C-3 60 431 398 8A 0 62 0.05 702.1
C-4 0 431 431 0% 0 0 0.00 0.0
only the WBL color source. As the TSS concentration increased the amount desorbed decreased
illustrating the importance of biosolids concentrations in the color removal mechanism. As TSS increases
color removal increases, but only to a certain point beyond which the reaction is shifted towards
desorption (i.e. too much TSS shifts the color concentration gradient towards desorption). Once calcium
is added in greater dosages, color removal percentages increase. For example, at wastewater treatment
plant TSS concentrations,Table 7 shows percent color removed increases with calcium dosages from 31%
to 46%to 56%. This percentage could be higher if the biosolids had a greater reserve capacity (i.e. lower
color loads on the biosolids). The largest percent color removed achieved in these experiments was 70%
(see Table 6). Under wastewater treatment plant conditions, 40 mg/l of soluble calcium and 2500 ppm
TSS, WBL color removal efficiency maximized at 59%.
The fourth adsorption isotherm experiment was a culmination of the results found in the first
three experiments. None of the three initial experiments were able to produce data that would generate
23

the desired adsorption isotherm found in Weber (1972), Eckenfelder (1966) and Stumm and Morgan
(1981). By varying the initial color concentrations of the reactors and subjecting them to wastewater
Table 7: Adsorption and Calcium Results from 3rd Adsorption Isotherm Experiment
Initial Color of Reactors is 414 scu
Ca Dose Color after % color removed Color adsorbed (mg) Soluble Ca TSS
Reactor m 6.5 hrs after 6.5 hrs pergramTSS m after 6.5 hrs
1 0 609 -47% -199 19 0.98
2 0 456 -10% -23 23 1.79
3 0 456 -10% -26 23 1.63
4 0 431 4% -5 24 3.17
5 40 464 -12% 43 55 1.17
6 40 367 11% 25 56 1.89
7 40 292 29% 57 56 2.14
8 40 285 31% 42 55 3.05
9 60 398 4% 15 77 1.06
10 60 306 26% 75 74 1.44
11 60 359 13% 44 75 1.26
12 60 222 46% 64 74 3.00
13 80 367 11% 49 92 0.96
14 AO
215 48% 106 92 1.87
15 195 53% 96 93 2.27
16 182 56% 78 98 2.99
C-1414 0% 0 0 0.00
C-2414 0% 0 0 0.02
treatment plant conditions and varying soluble calcium concentrations, the desired adsorption isotherm
was generated with both an adsorption limb and the beginning of the desorption limb. Figure 3 is an
illustration of the generated isotherms. The curves show that as calcium concentrations increase the color
removal mechanism is shifted towards greater adsorption. This can be seen in the upward and leftward
shifts in the curves. Note that the biosolids used in the 60 and 80 mg/I trial resulted in an increase in
reactor color of 28%. Thus the isotherms are conservative in their representation and actually the
biosolids remove a greater percentage of color than depicted. Table 7 also shows that soluble calcium
concentrations after 6.5 hours are greater than initial dosed concentrations. This may indicate an
equilibrium shift as insoluble calcium originally bound to the biosolids could be going into solution at the
same time that soluble calcium and color bodies are being adsorbed. An EDTA desorption of the biosolids
was conducted to determine the total capacity of the biosolids to remove color, the results of which are
24

listed in Table 8. The data suggest a maximum capacity for color removal by the biosolids of
approximately 150 mg of color per gram of TSS.
FIGURE 3: Adsorption Isotherm Curves for Calcium Dosages of 40, 60 and 80 mg/l
90.00
80.00
e
70.00
•
E rn 60.00
® °
a `0 50.00 e °
`o
Q m 40.00 °
O N
tj a30.00 P
•Ca dose=40 mgrl(6/28195)
20.00 • ■ Ca dose=60 mgA (8/1/95)
° •Ca dose=80 mcyl(82195)
10.00
0.00
0 50 100 150 200 250 300 350 400 450
Measured Final Color mgA pt-co units
TABLE 8: Results of EDTA desorption of biosolids
to determine their maximum capacity for color removal
Color(mg)Desorbed Color(mg) Adsorbed Maximum capacity of biosolids to
Date: per gram of TSS per gram of TSS remove color(mg color/gram TSS)
6/28/95 74.9 71.8 146.7
8/l/95 44.2 60.3 104.5
8/2/95 63.7 81.9 145.6
The data for the fourth adsorption isotherm are summarized in Appendix A. This data is the
basis for a linear regression analysis performed on the color concentration data to determine if a linear
relationship exists between initial and final color concentrations. Figure 4 is the result of the linear
regression analysis conducted. The regression statistics for each calcium dosage arc listed in Table 9.
The slope coefficients for each line are judged to be statistically significant from zero based on the p-
25

values, even at an alpha level as low as 0.001. Note that an alpha level of 0.05 is typically used to
determine statistical significance(Hamilton, 1992). Thus the extremely low p-values for the slopes
FIGURE 4: Linear Regression Model Curves
for the Fourth Adsorption Isotherm Experiment
450
400 • Measured 40( mgn)
-o-350 Predicted(40 mgn)
x Measured(60 mgn)
c 300 b--Predicted(60 mgn)
z
0 x Measured(80 mg/I)
o. 250 —Predicted(80 mgn) x
a
E
o` 200
o • :
U
n
LL
150 x
•
x
100 :
50 x
0
0.0 100.0 200.0 300.0 400.0 500.0 600.0
Initial Color mgfl ptco units
reinforces the statistical significance of these coefficients. Also, the adjusted R-squared values support
that the statistical relationship is in fact linear.
TABLE 9: Summary Statistics for Linear Regression Models
Calcium Slope Intercept Adjusted P-value
Dose(mg1l) coefficient coefficient R-squared (slope)
40 0.7337 -52.54 0.94 0.00019
60 0.643 21.27 0.95 0.00012
80 0.5601 -11.61 0.96 0.00007
The overall result from these linear regression models is that over the initial color range of 100-
400 scu the color removal efficiency is statistically constant. Above this color range it appears that the
adsorption mechanism is shifted towards desorption and color removal efficiency decreases. This color
removal efficiency is increased lvith increasing levels of soluble calcium in solution. Figure 5 summarizes
26

the adsorption isotherm experiments results on calcium dosages vs. percent color removed. Note that
these values incorporate the reserve capacity of the biosolids for color removal in order to produce a "true"
percent color removed at each calcium level.
FIGURE 5: Soluble Calcium Dosage (mgA) vs. % Color Removal
70%
60%
50%
i
40%
%Color Removed
30%
20%
10%
0%
40 60 80
Calcium Dosage mg/I added as CaC12 solution
3.1.2 Conclusions
As previously discussed color, biosolids and soluble calcium concentrations have all been
determined to be statistically significant in determining color removal. The adsorption isotherm
experiments set out to quantify the effects of varying the concentration of each on color removal. There
appears to be a range of concentrations that are optimal for maximizing the percent color removed for
both TSS and biosolids concentrations. TSS concentrations currently present in the wastewater treatment
plant are within the optimal range. Decreasing TSS concentrations (1000-2000 ppm) result in deficient
amounts of biosolids to remove color while high TSS concentrations (greater than 3000 ppm) shifts the
equilibrium towards desorption unless compensated by increased calcium dosages. It is obvious from all
four experiments that increasing the calcium concentration in solution increases color removal. Calcium
concentrations are critical to driving this adsorption mechanism. However the relative increase in
27

improved color removal as calcium increases must be balanced with economical and technical
considerations at the mill. A maximum of 70%of brown color was achieved. However, under wastewater
treatment plant conditions 59% was achieved which is still much higher than previously predicted. The
maximum capacity for color removal of the biosolids used in these experiments was calculated to be
approximately 150 mg color per gram of TSS. The proper balance in wastewater treatment plant
conditions could possibly be determined based on the ranges tested in the adsorption isotherm
experiments. Finally, it was determined using linear regression models that within the 100-400 scu initial
color range, the color removal efficiency is constant. Above this range it appears that the adsorption
equilibrium is shifted towards desorption. This clarifies an assumption provided by McCord (1995) that
shifts in initial color levels may result in more desorption.
3.2 Calcium Source Experiment
3.2.1 Results and Discussion
The results of the calcium source experiments provide insight into the differences between the
various sources of calcium tested as well as the effect of temperature on the color removal process. Table
10 summarizes the data from the four calcium sources tested at room temperature. Once again it is
important to consider the reserve capacity of the biosolids when comparing the color removed by each
source. The two solid forms of calcium, PCC and lime mud, both have low solubility and remove little, if
any, color. The color removed by PCC and CaCIZ would be better quantified if a WBL+biosolid control
would have been tested. The temperature experiment, discussed later in this section, will clarify the
assumptions made here. After considering the reserve capacity of the biosolids, both solid forms appear
not to contribute to the color removal process. On the other hand the two soluble sources of calcium,
CaC12 and Hardwood D1 filtrate, contribute much more to color removal. CaCl2 is an extremely effective
source of calcium with percent color removed ranging from 45% at the 25 mg/1 dose to 63% at the 55
mg/I dose. The percent color removed for the Hardwood D1 filtrates had to be adjusted volume
28

proportionally based on percent brown color. Recall from McCord (1995) that the color removal
mechanism only acts on brown color sources.
TABLE 10: Adsorption and Calcium Results from Original Calcium Source Experiment
Calcium Reactor PH of Ca dose male or % Color -�R Soluble Ca
Source 11 WBL m f mPA R emoval M?A
PCC « 1 7 25 C 259 22% 23
6/20/95
262 21%
2 8 25 23
40
3 2I 11 1 257 23% 24 1'.
4 7.5 40,
I I - 255 23% 24
M
5 7 55 236 29% 26
a 6 8 55 255 23% 26
CaC12 control 7.5 40
137 59% 47
CaC12 control 7.5 40 t 134 60% 46
WBI,solution 7.7 0 0%
11. 0
332
Calc*um 1 7 2l 170 49% 38
came
6/21/95
2 8 25 i 181 45% 37
3 7.5 40 139 58% 51
N
4 7.5 40 142 57% -N
51
5 7 55 122 63% 63
6 8 55 128 61% 63
YML solution 7.5 0 332 ii�l:
0% 0
Hardwood 7 1 7 25 45
216 42%
D1
6/22/95 2 8 25 214 42% 44
3 7.5 40 249 35% 62
4 7.5 40 7 35% 63
5 7 55 275 13% 83
6 8 55 285 10% 80
CaC12 control 7.5 40 138 61% 49
CaC12 control 7.5 40 i 139 60% 49
BL+Biosolids 7.5 0 333 5% 18
% BL+CaC12 7.5 40
337 4% 43
VvIBL solution 7.5 0 351 A
Lime Mud
1 7 25 202 of 18
6/23/95 y 2 8 25 235 33% 22
3 7.5 40 231 a34%
4 7.5 40 233 34% 21
213 39%
5 7 55 23
6 8 55 i . 215 ,A
23
CaC12 control 7.5 40 7 130 63% 48
CaC12 control 7.5 40 i 132 62% 49
BL+Biosolids 7.5 0 208 41% 19
BL+CaC12 7.5 40 329 6% 43
WBL solution 7.5 0
0
29

Therefore the full percent color removed in Table 10 is masked by the proportion of bleach color (the
source of calcium) in the reactors. This is also true for the full scale wastewater treatment plant percent
color removal. Table 11 corrects for this effect and gives the percent brown color removed with
Hardwood D1 as the calcium source. In this respect Hardwood D1 is as effective as CaCIZ in color
removal. At wastewater treatment plant calcium conditions (i.e. 40 mg/1 dose) percent color removed is
upwards of 60%. These results are in the same range as those determined in the adsorption isotherm
experiments.
TABLE 11: Corrected Values of%Brown Color Removed
by Hardwood D1 Calcium Source
Ca Dose Original %brown Original Color(scu) %Brown
mg/l Color color brown color removed color removed
25 370 75% 277.5 154 55%
25 370 750/- 277.5 156 56%
40 384 600/. 230.4 135 59%
40 384 60% 230.4 135 59%
55 315 45% 141.75 40 28%
55 315 45% 144 . 5 30 21 0
Note that all the CaC12 controls were designed at wastewater treatment plant conditions (40 mg/1
dose of Ca and pH = 7.5). A statistical analysis was performed on the color and soluble calcium
concentrations of the control reactors to test the distributions of these values (see Appendix B). Based on
the mean, median, standard deviation and skewness it was concluded that the control data was
approximately normally distributed thus statistically strengthening the results. For detailed descriptions
of the statistical parameters generated the reader is encouraged to consult Hamilton (1992). Reactor pH
did not have a significant impact in determining color removal. However, there was little difference in
the amount of HCI and/or NaOH added to the reactors to adjust the pH so conclusions based on pH
probably would not be proper in this circumstance.
Soluble calcium concentrations from Table 10 also show that the soluble sources of calcium
provide enough soluble calcium in solution to drive the color removal mechanism towards adsorption.
The solid sources result in soluble calcium concentrations no greater than 26 mg/l. This has been shown
30

to be too low to keep the mechanism shifted towards adsorption. Figure 6 compares the percent brown
color removed among the four sources at wastewater treatment plant calcium conditions. These values
have been corrected for biosolid reserve capacity.
FIGURE 6: % Brown Color Removed for each Calcium Source
at Centerpoint Conditions
70%
so%
so%
% BL color 40%
Removed
30%
20%
10%
0%
PCC CaC12 Hwd 01 Lime mud
Calcium Source
The second experiment tested the effects of temperature on color removal for a soluble source
(CaC12) and a low solubility source (PCC). Appendix C contains the results from this temperature effect
experiment. At first the temperature affect is hidden by the influence of the biosolid reserve capacity.
Figure 7 rectifies this problem and the effect of temperature on color removal is quite evident. The
conditions for Figure 7 are the wastewater treatment plant conditions previously mentioned. An increase
in temperature increases the percent color removed for both PCC and CaC12. The increase in color
removed for PCC can be attributed to the property of CaCO3 whereby its solubility increases with
temperature (Stumm and Morgan, 1981). Also note that increases in temperature increase desorption off
of the biosolids. Why is more color removed at higher temperatures when simultaneously more color is
being desorbed off the biosolids? A possible explanation could be that by desorbing more color off the
31

biosolids, the higher temperatures prepare the surface of the biosolids for a more efficient ordering of the
color bodies on the cell walls. If so, then the color removal values from the previous experiments may
actually underestimate the possible color removed at the wastewater treatment plant.
Figure 7: Temperature Effects on Color Removal for the Calcium Source Experiment
60% ......................................................_.............................................................................._._.............
,
50%
40%
30%
20%
%Color Removed 10%
0%
-10%
-20%
-30%
-40% ............................................_.-.......................................................................................................
i
CaC12 CaC12 PCC Biosolids Biosolids
629 7/11 629 629 7/11
❑ 26.5 degrees C Q 35 degrees C
3.2.2 Conclusions
Several important conclusions can be drawn from this area of research. First, only the soluble
sources of calcium (CaC12 and Hardwood D1) serve as effective sources for the color removal process.
The solid, low solubility forms contribute little to the process, even at higher temperatures. The
contribution of Hardwood DI as a viable source of soluble calcium is important in relation to the BFRTm
project. The calcium levels tested bracket the projected BFRTM conditions, thus lending support that
Hardwood D1 filtrate will provide a viable source and amount of soluble calcium needed for the color
removal mechanism to occur effectively while Pine filtrates are recycled by the BFRTM demonstration
project. An increase in temperature resulted in greater color removal percentages. This implies that the
color removal values produced by previous experiments may actually underestimate the potential color
removable at wastewater treatment plant conditions. Higher temperatures increase the desorption of color
32

off of the biosolids. Although this may seem negative, it appears that this process results in a rearranging
of color bodies on the cell walls of the biosolids and increases the percent color removed at the end of the
experiment. Researching the physical mechanism by which calcium and color are exchanged with the
biosolids in solution may lead to a better understanding of the color removal mechanism.
3.3 Sewer Generated Color Experiment
3.3.1 Results and Discussion
The results of the sewer generated color experiments apply to the issues of quantifying the
amount of sewer generated color on Pine and Hardwood DI filtrates and whether this sewer generated
bleach color is removable across the wastewater treatment plant. Previous work by Koelsch (1995) has
estimated 30% sewer generated color on Pine DI filtrates and up to 100% sewer generated color on
Hardwood D1 filtrates. McCord (1995), Caron (February 1992) and Koelsch (1995) have estimated 20%
sewer generated color at the sum of the sewers, which is at the influent to the wastewater treatment plant.
The results obtained in this experiment complement these previous findings. Table 12 shows that Pine DI
sewer generated color ranged between 26% and 47% while Hardwood DI sewer generated color ranged
between 59%and 101%.
TABLE 12: Results of Sewer Generated Color on Pine and Hardwood D1 samples
Sample: Date: % S.G.C.
Pine D1 617/95 45%
Pine D1 6/8/95 26%
Pine D1 7112195 47%
Pine D1 7/13195 32%
Hwd D1 6/7/95 101%
Hwd D1 6/8/95 59%
Hwd D1 7/12/95 64%
Hwd D1 7113195 88%
33

The second stage of the sewer generated color experiments was to submit the filtrate blends to
wastewater treatment plant conditions. All the data collected for this experiment can be found in
Appendix D while Table 13 lists the color before SGC conditions, after SGC conditions and after
wastewater treatment plant processing. On 7/12/95, there was a net decrease in hardwood filtrate color
TABLE 13: Results of Sewer Generated Color Filtrates Subjected to WTP Conditions
Net color
Color Color Color increase
before after after across
Sam le ID Sam le Date S.G.C. S.G.C. WTP WTP system
Pine(1) 7/12/95-- 1172- 1695— 1 116 12%
Pine(2) 7/12/95 1157 1647 1453 26%
Hwd(1) 7/12/95 591 733 516 -13%
Hwd (2) 7/12/95 596 710 507 -15%
Bleach(1) 7/12/95 921 1280 986 7%
Bleach(2) 7/12/95 921 1242 1041 13%
Pine(1) 7/13/95 613 791 732 19%
Pine(2) 7113/95 594 1004 742 25%
Hwd(1) 7/13/95 511 688 562 10%
Hwd(2) 7/13/95 493 685 552 12%
Bleach (1) 7/13/95 569 748 679 19%
Bleach(2) 7/13/95 569 867 658 16%
across the wastewater treatment plant which indicates that for that day all of the sewer generated color
and some of the original bleach color was removed across the wastewater treatment plant. This result was
not duplicated the following day. The important point from this data is that there exists a possibility that
hardwood bleach color may be removable across the wastewater treatment plant as well as sewer
generated color. If this is the case then this is excellent information for BFRTm. With the pine filtrates
being recycled, the possibility of hardwood filtrate bleach color being removable adds to the effectiveness
of BFRTM. However, the fact that only two trials were run and unable to be duplicated makes this idea
unsubstantiated pending further research. The remainder of the data shows a net increase in color across
the wastewater treatment plant as expected from results by McCord (1995). Comparing the color after
34

SGC conditions and color after wastewater treatment plant indicate that a portion of SGC is removable
across the wastewater treatment plant.
3.3.2 Conclusions
This experiment was successful in complementing previous work on SGC of Pine and Hardwood
D1 filtrates, but also raised questions on the possibility of color removal of regular and sewer generated
hardwood bleach color. Sewer generated color for pine D1 filtrates was 32-47% and for hardwood D1
was 64-88% which agrees with previously collected data. A portion of SGC for all the bleach filtrate
blends is removable. However pine and combined bleach flows experienced a net color increase across the
sewer and wastewater treatment plant systems. The data did suggest that some Hardwood bleach color
may be removable across the above systems. The data was unable to be duplicated over the two day
period,but further research would substantiate or refute hardwood bleach color removal in the wastewater
treatment plant.
3.4 BFRTM Simulation Experiment
3.4.1 Results and Discussion
The BFRTM simulation experiments generated data for both the short-term (i.e., directly after
BFRT becomes on-line) and long-term (i.e., after the system has equilibrated to lower color loads)
scenarios. Trials 2 and 4 mimic short-tens conditions while Trials 1 and 3 mimic long-term conditions.
Tables 14, 15 and 16 list the color removal and soluble calcium data from all four trials. One must
consider the four variables (calcium dose, initial color level, blend ratio and biosolids concentration) in
this designed experiment when interpreting the data. Also note that the calcium level targeted for each
reactor was not met. It was intended that the calcium levels set forth in the designed experiment would be
the actual calcium levels in the reactors. The reactors were instead dosed with the listed concentrations of
soluble calcium, but the concentration of calcium in the hardwood Dl filtrate was not taken into
consideration. Based on the six month average hardwood D1 calcium concentration, 167 mg1l, the actual
35

calcium concentrations of each reactor were corrected in Tables 14 and 15. These approximations are
based on the assumption that the hardwood and dosed calcium concentrations are additive and fully
present in soluble form available for the color removal mechanism. The Mettler titrator was unable to
36

Champion International Corporation - Canton, NC
Table 14: BFRTM Simulation Experiment - week of 7/17/95
Calcium dose Corrected Ca Initial Blend Ratio Biosolids Trial#1 *-Biosolids Desorbed Trial#2 **-Biosolids not Desorbed
mg/1 added level (mg/I) Color % Brown cone. Final % Color Final % Color
Jar# Dav# as CaC12 with Hwd DI mg1I Color (m ) Color Removed Color Removed
1 1 45 57 380 85 2500 140 63% 149 61%
2 1 45 57 380 85 2000 158 58% 151 60%
3 1 45 79.6 380 55 2500 183 52% 191 50%
4 1 45 79.6 380 55 2000 215 43% 190 50%
5 1 45 51.3 200 85 2500 80 60% 89 56%
6 1 45 51.3 200 85 2000 89 56% 85 58%
7 1 45 63.2 200 55 2500 97 52% 109 46%
8 1 45 63.2 200 55 2000 125 38% 93 54%
9 1 35 55.8 290 , 70 2250 148 49% 149 49%
10 1 35 55.8 290 ; 70 2250 146 50% 144 50%
1 2 25 37 380 85 2500 169 56% 275 28%
2 2 25 37 3801 85 2000 222 42% 286 25%
3 2 25 59.6 380 i 55 2500 215 43% 273 28%
4 2 25 59.6 380i 55 2000 235 38% 257 32%
5 2 25 31.3 200 , 85 2500 106 47% 153 24%
6 2 25 31.3 200 ! 85 2000 106 47% 160 20%
7 2 25 43.2 200 ; 55 2500 124 38% 168 16%
8 2 25 43.2 200 , 55 2000 156 22% 153 24%
9 2 35 55.8 290f 70 2250 156 46% 220 24%
10 2 35 55.8 290 70 2250 156 46% 214 26%
* Color was desorbed off of biosolids using deionized water and aeration prior to centrifugation
** Biosolids mimic those in current aeration basin system

w • `
a o
W O O b 00 J D\ U A W N ► �p 00 �1 U A W N r 'wt cr
y O^ N
< N
N. d y N N N N N N N N N N r r r r r r r r r rtz
Y
0.
. o
n
O O W W N N N N N N N N W W 0 C _V]
a m
0
^ r o'
2• io 0 7
Urr n K
w F e e oo m e e W w
(.
r o ❑ �
vP N' N N N N N N W W W W N N N N N N
O w OD 00 00 �O �O O O O O 00 00 00 00 3 O .n SV
O O O O O O O O O O O O O O O O 0 0 0 0 O
W O (D
n
a �1 J U U o0 0o U U Go 0o J J U U o0 00 U U 00 00 O W G �
7 �
v � N
�• N N O U O U O U O U N N O U O U O U O U � O y
n U U O O O O 0 0 0 0 U U 0 0 0 0 O O O O O O O
� y N
n _
a � n
C N N �-• N N W W N N r N N N N n •� O
O N W O O W W w O O p � a
U T O� �l A A O W O O J N U O U A W � A 0 0 y
c. ^ O
O 03
p x
eo O
O
N
o e e e o e o o e o e o 0 o
S �
n -
a Z
y �
N N r r W N W W N N r N N N N n
W w U O\ O� N b a1 W O� A N N O r U U O� O S.
EL
00 N O\ �"' O\ A 00 a �-+ U N W N J A 00 A 00 00 O
't
x
x
O
r
Oo O N O
o e e o o laz
n e C
N
0
R
A
a

Champion International Corporation - Canton, NC
Table 16: BFRT^' Simulation Experiment Soluble Calcium Data Measured after 6.5 hours
Calcium level Initial Color Blend Ratio Biosolids Trial#1 * Trial#2 "" Trial#3 " Trial#4**
Jar# Da # m I as CaC12 mg/1 % Brown Color cone. (m ) mg/1 m m m
PA
I 1 45 380 85 2500 59 46 44 43
2 1 45 380 85 2000 53 41 47 47
3 l 45 380 55 2500 71 65 59 60
4 1 45 380 55 2000 72 64 60 61
5 1 45 200 85 2500 55 43 43 45
6 1 45 200 85 2000 53 44 44 46
7 1 45 200 55 2500 56 44 48 52
8 1 45 200 55 2000 60 58 52 52
9 1 35 290 70 2250 42 37 42 35
10 1 35 290 70 2250 38 42 42 32
1 2 25 380 85 2500 26 27 29 28
2 2 25 380 85 2000 32 32 29 28
3 2 25 380 55 2500 49 26 40 41
4 2 25 380 55 2000 50 51 39 42
5 2 25 200 85 2500 26 2 30 24
6 2 25 200 85 2000 28 31 26 25
7 2 25 200 55 2500 37 18 32 30
8 2 25 200 55 2000 36 40 36 36
9 2 35 290 70 2250 42 32 49 42
10 2 35 290 70 2250 45 45 45 41
* Color was desorbed off of biosolids using deionized water and aeration prior to centrifugation
"*Biosolids mimic those in current aeration basin system

as well as the 60 and 80 mgp calcium dosage trials for the fourth adsorption isotherm experiment. The
data clearly illustrate that the color bodies are physically being removed from solution in accordance with
previous findings by McCord(1995).
3.4.2 Conclusions
Although the experiment conducted did not completely simulate projected BFRT conditions,
valuable conclusions can still be drawn from this research. By accounting for the calcium concentration
of the hardwood D 1 in each reactor, several reactors (1,2 and 5-8) do fall within the range of projected
BFRT" conditions. These reactors show that color removal efficiencies between 42-56% were achievable
at the high color, high blend ratio combination and between 22-38%for the low initial color, low blend
ratio situation. Recall that a high blend ratio contains a high percentage of brown color compared to
bleach color while a low blend ratio has a high percentage of bleach color compared to brown color.
These color removal efficiencies are,for the most part, above the current 30%efficiency at the wastewater
treatment plant measured on combined bleach/brown color. By increasing the soluble calcium
concentration (i.e. through adding soluble calcium to the upper basin of the wastewater treatment plant),
color removal efficiencies up to 63%were achieved. This option, however, has technical and economical
considerations beyond the scope of this report. Finally, color removal efficiencies were greater for the
long-term conditions than for the short-term conditions. This suggests that it will take time for the
biosolids to equilibrate to lower color loads, but once this is accomplished color removal efficiencies may
be greater than what initially occurs as BFRTM goes on-line at the mill.
41

4. Research Conclusions: Implications of BFR
Although each area of experimentation was created to answer specific questions of interest to
Champion, several general conclusions can be made and applied to the implications of BFRTM operating
at the Canton, North Carolina mill. The adsorption isotherm, calcium source, sewer generated color and
BFRTM simulation experiments all provide insight for the possibilities of this technology for color removal
as well as better characterizing the processes that are currently occurring. These BFRTM implications are
as follows:
1) Calcium is the driving mechanism for the color removal mechanism occurring at the
wastewater treatment plant. As long as calcium concentrations remain above 3540 mg/l, lower color
concentrations should not adversely affect color removal efficiencies. McCord (1995) raised this concern,
but according to the linear models created in this report there is little need to be concerned unless initial
color load concentrations increase above 400 scu.
2) Hardwood D1 filtrate serves as a suitable source of calcium at brown color to bleach color
ratios greater than 55%. This ratio represents a lower bound for the projected conditions under BFRT .
Therefore anticipated post-BFRTM conditions should present a blend ratio higher than 55%. The color
removal efficiency after BFRTM is a function of calcium concentration, initial color concentration, blend
ratio and biosolid concentration. What is important is that the key component of the mechanism, calcium
concentration, will be satisfied by the Hardwood DI filtrate which will continue to be sewered to the
wastewater treatment plant. UnderBFRTM at the Canton mill, only pine bleach filtrates will be recycled.
3) There exists a possibility for the removal of both sewer generated and regular bleach color for
Hardwood filtrates although more extensive testing would be required to substantiate this possibility. This
is beneficial because it could mean increased percent color removal across the wastewater treatment plant.
This could be an unexpected removal of color beyond that of removing the contribution of sewering pine
filtrates due to BFRTM operating.
42

4) A decrease in sewer generated color due to the removal of Pine filtrates should also result in
less unaccountable color throughout the mill. This would improve the quantifiability of the mill's current
color balance database.
5) Current wastewater treatment plant operating conditions appear reasonably optimum for color
removal. The average biosolids concentration, MCRT, hydraulic residence time and soluble calcium
concentrations of the wastewater treatment plant used throughout this project typically resulted in
optimum color removal efficiencies. Therefore no apparent changes to the operation of the wastewater
treatment plant seem evident.
Based on the above research, BFRTM, if it is successfully demonstrated, may assist Champion's
continued efforts to operate below the NPDES permit level for color discharges. Each area of
experimentation adds information to various components of treating the wastewater generated by the mill
as well as contributing to the primary objective of determining BFRTM possibilities. This report is
intended to be a valuable resource to Champion International Corporation as well as the pulp and paper
industry as a whole. The BFRTM process may significantly enhance the paper industry's ability to produce
quality paper products in both economically and environmentally efficient manners. A recent article on
Champion stated the significance of this process best:
"The hope is that BFRT"r will offer the U.S. paper industry the ability to modify existing mills at
considerably less cost than building new,environmentally comparable facilities(Swann, 1995)"
43

5. Suggestions for Further Study
Throughout the course of this project questions arose beyond those which were addressed in this
report. The following are a list of possible experiments that could be conducted by Champion in the near
future as a continuing effort to enhance the efficiency of the color removal mechanism occurring across
the wastewater treatment plant.
1) A full scale trial of calcium additions across the wastewater treatment plant could be
performed to determine the full effects of calcium dosages on color removal. Soluble calcium would be
added to the upper aeration basin of the wastewater treatment plant in various concentrations, such as
those used in this project, and color measured on secondary effluent samples periodically. Both the level
of color removal and economic and technical feasibility of such calcium additions could be quantified.
However, it appears that optimum conditions for color removal have been achieved.
2) The determination of the roles of other metals(for example magnesium) on color removal was
a suggestion for further study raised by McCord (1995). Due to the intense nature of the experiments
conducted, there was not enough time to address this issue. The advent of BFRTm substantiates
performing such analyses. The recycling of pine filtrates may alter the availability and forms of metals
currently present in the sewers.
3) The results of the sewer generated color experiments imply that continued testing of whether
or not sewer generated and regular bleach color of hardwood filtrates is truly removable is important.
Determining whether or not the data in this report is an anomaly is significant in quantifying maximum
color removal possibilities under BFRT . Also the mill's current color balance could be better
characterized (i.e. less unaccountable color).
44

4) During the temperature effect calcium source experiments, it was concluded from a step-wise
regression that the volume of calcium added was the most significant variable in determining color
removal. The only factor accounting for such volume differences was the size of the reactors used. The
calcium concentrations added were the same for both types of reactors used. The jar stirrers used 1000
milliliter reactors while the waterhath used 750 milliliter reactors and the temperature was held constant.
Thus the only difference was that the waterbath reactors were aerated while the jar stirrers were physically
stirred. Thus an experiment could be conducted to determine how different types of mixing (oxygen
aeration, nitrogen aeration and physical stirring) affect color removal. The results of such an experiment
could lead to technical renovations in the aeration basin that could increase color removal.
5) A final experiment could be conducted to trace the path of calcium across the wastewater
treatment plant in order to better understand the role of calcium in the color removal mechanism. An
option may be to isotopically label calcium ions if this is possible. This experiment, if possible, would
result in a mechanistic model describing the movement of calcium throughout the process.
45

6. Conclusion
The research presented here is the most recent in a long line of projects conducted by Champion
since the early 1980's to find solutions to the environmental problems faced by the pulp and paper
industry. Due to the environmental sensitivity of its location, the Canton mill has been the site of quite
significant renovations such as the Canton Modernization Project and the current installation of the
BFRTM demonstration project. Such technological improvements have and will continue to result in
producing high quality paper in both economically and environmentally friendly manners. The BFRTm
technology, if constructed and successfully operated, may lead the industry closer to a closed mill system.
.The efforts of this report show early promise for this-revolutionary technology. The results presented
here will hopefully benefit the pulp and paper industry as a whole, as well as wastewater treatment plants
across the country.
7. Acknowledgments
I wish to thank Dr. Gabriel Katul of Duke University, Susanne Koelsch, Derric Brown, Bob
Williams and many other members of the Champion International EOHS staff. I would also like to thank
those at the wastewater treatment facility and summer intern Dallas Gamble. Without their guidance and
support this project would not have been possible. A special thanks is extended to Susanne for her
tremendous assistance and guidance.
46

APPENDIX A:
Summary of Results from the Fourth Adsorption Isotherm Experiment
g. Final Percent Color rem. Final Percent
Ca dose Initial
mg/I added Color 2 Color Color (mg) per Soluble Ca Calcium
Date Reactor as CaC12 mg/I mg1I removal gram TSS mg/I Recovered
6/28/95 1 40 141.3 j 73 48% as 53.38 90%
6/28/95 2 40 211.9 112 47% 31.6 g 52.33 as
6128195 3 40 282.5 § 142 50 47.0 Z 50.87 855
40 353.1
6/28195 4 191 46% 58.7 50.81 85%
6/28/95 5 40 423.8 222 48% 64.9 51.60 87%
6/28/95 6 40 494.4 304 39% 71.8 49.83 84%
6/28/95 7 40 565.0 y 402 29% 65.5 § 43.75 73%
0 565.0 562 %1 1.1
6/28/95 Control-1 19.61
8/1/95 1 60 100.0 .A 60 40% 16.7 68.97 90%
8/1/95 2 60 150.0 75 50% 30.7 64.65 85%
8/1/95 3 60 200.0 97 525 43.3 66.25 87%
8/1/95 4 60 250.0 128 49% 52.1 65.35 86%
811/95 5 60 300.0 154 49% 60.3 65.50 86%
8/1/95 6 60 350.0 "N' 208 41% 59.2 64.48 84%
13/1/95 7 60 386.0 245 37% 59.5 64.05 84%
-1 8/1/95 Control 0 386.0 @ 496 -28% -46.6 16.31
8/2/95 1 80 100.0 48 52% 22.5 88.06 965
812/95 2 80 150.0 5 9 57% 36.4 89.71 97%
8/2/95 3 80 200.0 86 57% 49.1 85.32 93%
4 8/2/95 80 250.0 05 58% rye 82.69 90%
95 80 300.0 1 1
59% 76.9 82.66 90%
8/2/95 6 80 350.0 165 53% 81.9 81.94 89%
7 80 386.0
8/2/95 196 49% 81.9 80.84 88%
812/95 Control-I 0 386.0 496 at -49.5 12.16
47

APPENDIX B:
Statistical Analysis of Final Color and Calcium Concentration Data for Control
Reactors Used in the Calcium Source Experiment
Color summary statistics Calcium summary statistics
Mean 136.375 Mean 48.623
Standard Error 1.426002154 Standard Error 0.6886429
Median 137.5 Median 48.568
Mode 139 Mode #N/A
Standard Deviation 4.033343172 Standard Deviation 1.9477762
Sample Variance 16.26785714 Sample Variance 3.793832
Kurtosis -0.815073242 Kurtosis -0.5925304
Skewness -0.399251505 Skewness -0.2203707
Range 12 Range 5.604
Minimum 130 Minimum 45.544
Maximum 142 Maximum 51.148
Sum 1091 Sum 388.984
Count 8 Count
8
Confidence Level(95.000%) 2.794908725 Confidence Level(95.0001/o) 1.3497132
48

APPENDIX C:
Summary of Results from Calcium Source Experiment Testing the Effect of
Temperature on the Adsorption Process
Room Temperature=80 degrees F '! Waterbath Temperature=95 degrees F
Calcium pH of Ca dose Final Color % Color %Calcium Final Color % Color %Calcium
Source Reactor WBL m MgA Removal Recovered m Removal Recovered
PCC 1 7 25 565 -8% 42% 555 -6% 28%
6/29/95 2 8 25 555 -6% 41% 545 4% 33%
3 7.5 40 f 532 -2% 32% 539 -3% 35%
4 7.5 40 549 -5% 31% 542 4% 22%
5 7 55 529 -1% 29% 507 3% 21%
6 8 55 545 4% 27% 578 -11% 27%
CaC12 7.5 40 368 30% 72% 1, 257 51% 88%
control-1
CaC12 7.5 40 362 31% 73% 281 46% 82%
control-2
Biosolid 7.5 0 552 -6% 652 -25%
w/o Ca
CaC12 1 7 25 233 26% 79% 303 4% 72%
7/11/95 2 8 25 227 28% 81% 290 8% 73%
3 7.5 40 189 40% 84% 191 39% 75%
4 7.5 40 209 33% 83% 238 24% 76%
5 7 55 169 46% 83% 202 36% 81%
G 8 55 172 45% 84% 191 39% 83%
Biosolid 7.5 0 305 3% 411 -31%
w/o Ca
49

APPENDIX D:
Summary of Results from the Sewer Generated Color Experiments
Net color
Color Color Color increase
before after after across
Sample ID Sample Date S.G.C. S.G.C. WTP WTP system
Pine(1) 7/12/95 1172 1695 1316 12%
Pine (2) 7/12/95 1157 1647 1453 26%
Hwd (1) 7/12195 591 733 516 -13%
Hwd (2) 7/12/95 596 710 507 -15%
Bleach (1) 7/12/95 921 1280 986 7%
Bleach (2) 7/12/95 921 1242 1041 13%
Pine(1) 7/13/95 613 791 732 19%
Pine (2) 7/13/95 594 1004 742 25%
Hwd (1) 7/13/95 511 688 562 10%
Hwd (2) 7/13/95 493 685 552 12%
Bleach (1) 7/13/95 569 748 679 19%
Bleach (2) 7/13/95 569 867 658 16%
Calcium Calcium Calcium Net Calcium
mg/l mg/I mg/I increase
before after after across
Sample ID Sample Date S.G.C. S.G.C. WTP WTP system
Pine 7/12/95 116 0 123 6%
Hwd 7/12195 101 0 0 -100%
Bleach 7/12/95 108 0 125 16%
Pine 7/13/95 87 63 0 -100%
Hwd 7/13/95 108 101 0 1 -100%
Bleach 7/13/95 96 78 123 28%
Color Color Calcium Calcium
before after mg/l before mg/l
Sample ID Sample Date S.G.C. @ S.G.C. S.G.C. @ after S.G.C. % S.G.C.
Pine D1 7/12/95 1162 1709 300 0 47%
Pine D1 7/13/95 590 776 227 169 32%
Hwd D1 7/12/95 359 590 0 0 64%
Hwd D1 7/13/95 329 619 0 0 88%
Pine Eo 7/12/95 1392 19
Pine Eo 7/13/95 1131 12
Hwd Eo 7/12/95 959 11
Hwd Eo 7/13/95 798 9
@ S.G.C. refers to the sewer generated color conditions
50

1
l
1 '
J C% U A W N n Q
to
-1 ^
O J T U A w N �- n N N N N N N N N N N N e r r �-- r r-• y
6 O
O b U A A In 1n �,;,� N 00 to b U A A U U �j �j 00 U � ~C• �.y
i. in W W J lJ
A A PO b W
.7 J J 00 J a\ U DD J D\ O� � 00 00 00 00 00 00 DD 00 00 Oo O N .7
C a e e o e e o o e e '" o e e o e e e e e e ~
r+ R ^ O
3 "C
to ,A„. o `oo `w .". b z •owe 3 � ❑• J J In U W W rn o. J J to to ��w
tz
ri
�.0.� NA
e
W rn rn J J J z J 3 H
9 h 5 C o
0 0 0 0 0 0 o o
0 o e o 0 0 3 Flo
W 7 .7
R �
v
r•
A J T A 3 G O O. W W O W �.., �p OO J O 01 W W O W �... �O 0o J ~
El r- Oo In 00 b A N J b U ::'F .... OO U m �D A N J ip U pj
n o F �hzlf
tD b N N 00 A O U ONi 41.
H
o e e \ e a 'ItO J W J O. O1 U J J O. O. O� 0, J .J-. D\ O. J J O� �•
-4 m b O U A W A A b
0 0 o N O A
a -�

REFERENCES
American Public Health Association. 1992. Standard Methods For the Examination of
Water and Wastewater. 18th edition. Washington D.C.
Caron,B. January 20, 1992. Sources of Sewer Generated Color. Memo.
Pensacola,Florida.
Caron,B. February 11, 1992. Cause and Control of Sewer Generated Color. Memo.
Pensacola,Florida.
Carpenter, W.L. March 14, 1986. Color Measurement Procedure. Notes to File. National Council of the
Paper Industry for Air and Stream Improvement, Inc. Gainesville, Florida.
Champion International Corporation. April 30, 1995. 1995 Color Removal Technology
Report. Canton,North Carolina.
Champion International Corporation. July 1995. Datastream. Database.
Eckenfelder, W.W. 1966. Industrial Water Pollution Control. McGraw-Hill,
New York,New York. pp. 100-109.
Hamilton,L.C. 1992. Regression with Graphics: A Second Course in Applied Statistics.
Duxbury Press, Belmont, California.
Henderson,B.,Hilleke,J.,Koelsch, S., Steinberg,M.and Brierley, C. February 16,
1995. "Canton Color Removal Project Planning Meeting 2/9/95." Memo.
Hilleke,J.,Henderson,B.,Koelsch, S., Steinberg,M.and Salisbury, C. July 13, 1995.
Color Teleconference. Canton,North Carolina.
Koelsch, S. K. September,23, 1994. Wastewater Treatment Plant Color Removal Mechanism Studies
Status Report. Canton,North Carolina.
Koelsch, S.K. July 7, 1995. Current Canton mill Color Balance. Memo. Canton,
North Carolina.
Koelsch, S.K. and Salisbury, C. M. July 7, 1995. Color Teleconference. Canton,
North Carolina.
Maples, G., Ambady,R, Caron,J.R, Stratton, S. and Canovas, R.V. 'BFR: A new
process toward bleach plant closure.",TAPPI. vol. 77, no. 11,p. 71, 1994.
McCord, A. 1994. A laboratory analysis of color removal across a pulp and paper mill
wastewater treatment facility, Canton,North Carolina. Masters project. Duke
University,Durham,North Carolina.
Mettler Instrument Corporation. 1993. Operation Instructions: Mettler Toledo
DL77/DL70ES/DL67 Titrators. Hightstown,New Jersey.
National Council of the Paper Industry for Air and Stream Improvement,Inc. December 30, 1971.
An Investigation of Improved Procedures for Measurement of Mill Effluent and Receiving
Water Color. Technical Bulletin Number 253. New York,New York.
52

Pryately,J. October 1995. Personal Communication. Canton,North Carolina.
Steinberg,M. June 13, 1995. Color Teleconference. Canton,North Carolina.
Stratton, S. "Canton Sodium Bleach Trial Part I-Color Load Impacts". Memo.
April 15, 1987.
Stumm W. and Morgan,J.J. 1981. Aquatic Chemistry: An Introduction Emphasizing
Chemical Equilibria in Natural Waters. 2nd edition. Wiley-Interscience,
New York,New York.
Swann, C. E. 1995. Company of the year. Papermaker. 58: 23-41.
Weber Jr, W.J. 1972. Physicochemical Processes for Water Quality Control. Wiley-
Interscience,New York,New York. pp. 204-211.
53

A LABORATORY ANALYSIS OF COLOR REMOVAL ACROSS A PULP AND PAPER MILL
WASTEWATER TREATMENT FACILITY
CANTON, NORTH CAROLINA
by
Aimee Winter McCord
Date: 6- n , �f
Approved:
Dr. a ' Katul Advisor
Dr. Norman L. Christensen, Dean
Master's Project submitted in partial fulfillment of the requirements for the Master of Environmental
Management degree in the School of the Environment of Duke University
1995

Abstract: The Champion International Corporation pulp and paper mill located in Canton, North Carolina
produces approximately 24 to 27 million gallons of wastewater per day. This wastewater is discharged into the
Pigeon River. Because of the small flow volume of[lie river, wastewater can account for a large percentage of the
river flow. Therefore, the mill's wastewater constituents are closely monitored. The mill is one of only a few mills
in the U.S. that must comply with an effluent color standard. Color at the mill is produced when wood chips are
brokendown and made into bleached pulp for paper. To gain a better understanding of color removal at the mill's
wastewater treatment plant,a series of studies was conducted to identify the factors affecting color removal. This
project is an attempt to provide background for these experiments, present results and draw several conclusions
from the work. Through these studies, it was found that color removal at the plant is dependent on calcium and
biosolids concentrations in the wastewater. Physical adsorption of color onto biosolids accounts for a significant
portion of color removal. Finally, it was determined that the color removal process acts on color from the pulping
operations in the mill but does not affect color produced during the mill's bleaching operations.

TABLE OF CONTENTS
I. Introduction I
II. Background 2
HA. Pigeon River Back rp ound 4
III. Sources of Color 10
IV. Historical Color Loads 14
V. Significance of Color 16
VA. Aesthetics 19
V B. Ecosystem Productivity 21
VI. Governing Policies and Permitting 24
VII. Wastewater Treatment 31
VIII. Obiectives of Research 34
IX. Areas of Study 34
X. Technigues of Color Measurement 36
XI. Buffer Experiments 37
XII. Desorption Tests 38
XII A. DesiV_n and Methods 38
MI B. Results and Discussion 42
XII C. Conclusions 47
XIII. Desorb-Adsorb-Desorb Experiments 48
XIII A. Methods and Design 48
X111 B. Desorb-Adsorb-Desorb Experiment al-Results and Discussion 51
"IC Conclusions for Desorb-Adsorb-Desorb Experiment 91 53
Xlll D. Desorb-Adsorb-Desorb Experiment k2 Methods and Design 54
XIII E. Conclusions from Desorb-Adsorb-Desorb Experiment N2 Sti
iii

XIII F. Desarb-Adsorb-Desorb Experiment 43 -Methods and Design 56
XI11 G. Desorb-Adsorb-Desorb Experiment #3-Results and Discussion 58
X111 H. Conclusions from Desorb-Adsorb-Desorb Experiment#3 59
XIII L Desorb-Adsorb-Desorb Experiment#4-Methods and Design 59
XIII J. Conclusions from Desorb-Adsorb-Desorb Experiment#4 61
XIV. Overall Conclusions 62
XV. Comparison of Results to Past Experiments 64
XVI. Current Studies 66
XVII. Suggestions for Further Study 66
XVII A. Calcium 66
XVII B. Bleach Filtrate RecvcleTA4 67
XVII C. Biosolids 69
XVIII. Future of lndustry 69
XIX. Conclusion 71
XX. Acknowledgments 71
Appendix A 72
Appendix B 73
Appendix C 76
References 78
iv

A
O
cC
C7
3G
U
Q
A
z
a
z
0
F
U
A
O
E•
z

I. Introduction
During the summer of 1994,the Environmental Department at Champion International
Corporation's Canton,North Carolina mill conducted a series of laboratory experiments to analyze
the factors that affect color removal at the mill's wastewater treatment plant. These experiments
are part of an ongoing effort by Champion to gain a better understanding of the color removal
processes at the wastewater treatment plant in order to maintain and improve color removal
efficiency.
-- —This project will-focus on the results from these experiments. First, the background
necessary to understand the significance of the results will be provided. This background section
furnishes general information regarding the mill and its surroundings, sources of color and the
regulations which affect the Canton mill's color discharge. Once the reader is familiar with the
issue;the objectives, experimental methods, results and conclusions from the summer work are
discussed. Finally, overall conclusions from the experiments are made, current efforts are
mentioned and suggestions for further study are offered.
The experiments which are presented here can be divided into three categories:
1) Buffer experiments that examine the effects of pH buffer on color measurements,
2) Desorption experiments that measure the ability to physically or chemically desorb color from
wastewater treatment plant biosolids,and
3) Desorb-adsorb-desorb experiments which measure the desorption of color off of biosolids, the
readsorption of color onto biosolids and, once again, the desorption of color off of biosolids.
These experiments lead to conclusions regarding the effects of buffer. color concentrations,
metals concentrations and biosolids concentrations on color removal and color measurements at the
Canton wastewater treatment plant. It is envisaged that the conclusions from these experiments
will serve not only the Canton mill but other wastewater treatment plants as well.

II. Backeround
A pulp and paper mill owned and operated by Champion International Corporation sits in
a rural valley in western North Carolina. Over the past 85 years the small town of Canton has
built up around this mill and survived on the business and jobs Champion brings to the region.
Canton is divided by the Pigeon River which runs through town and the mill. The Pigeon flows
from its origins in the Shining Rock Wilderness Area, across the Tennessee border, merges with
the French Broad River and empties into Douglas Lake. The impact on the river and the
surrounding ecosystems due to the effluent streams from the facility are of great concern to both
Champion and local residents. Although current technological advances have significantly
improved water quality, several factors still have the potential to affect the aesthetic and biological
conditions in the receiving waters. These effects are especially significant considering the
proximity of the mill to the state border. The nearby residents of Tennessee have, in the past,
raised serious concerns regarding the pollutants from across the state line.
Because of the high pollutant load from the mill (relative to the small size and pristine
nature of the Pigeon River) and the proximity to Tennessee, there are numerous federal and state
laws and regulations that govern the river quality and mill discharges. Some of these policies
include a National Pollutant Discharge Elimination System (NPDES) operating permit for the mill,
state water body classifications for the river, new proposed federal regulations governing air and
water emissions for the pulp and paper industry, and the basinwide plan for the French Broad
River Basin.
Historically, one of the issues of greatest concern regarding the Canton mill has been the
increased water color of the Pigeon River due to mill effluent. To minimize this impact, the
Canton mill is required in its NPDES discharge permit to meet stringent in-stream color
2

limitations. These standards, in turn, translate to reduced mill effluent color. Color is closely
monitored both at the mill and further downstream (approximately one-half mile past the North
Carolina/Tennessee border). In 1993 Champion completed a$330 million dollar Canton
Modernization Project. This project has helped Champion maintain effluent color loads well below
the required permit limit for the past several years.
When asked to define color, there is no one correct answer. The definition is likely to be
highly dependent on both the context of the question and the individual who is asked. Color
standards are normally defined narratively emphasizing that color is often in the eye of the
beholder. But, in the pulp and paper industry, how color is defined and measured can make a large
difference in the operating conditions of a mill. According to industry standards. color of pulping
waste is defined as "...the color of the light transmitted by the waste solution after removing the
suspended material, including the pseudocollodial particles" (NCASI, 1971). It is noted in the
definition that particles must be removed from a sample before water color can be measured. This
is necessary because wastewater color measurements can be significantly increased when
suspended materials that reflect light are present (NCASI, 1971).
To gain a better understanding of the color removal processes at the wastewater treatment
plant, Champion conducted a series of color removal experiments during the summer of 1994.
These experiments were an effort to better determine the factors that are most influential in the
process of color removal across the Canton mill wastewater treatment facility. The studies
consisted of benchscale reactors aerated under a varietv of conditions. Conditions varied as a
function of color source, metals concentrations. biosolids concentrations and several other factors.
3

M. Pireon River Background
As seen in Figures I and 11,the Pigeon River originates in the mountains of Haywood
County above Canton at an elevation of over 5,000 feet and is a part of subbasin 05 in the French
Broad River Basin (Clark, 1993). From its origins, the east and west branches of the river join and
flow in a northwesterly direction into Tennessee. The Pigeon continues on by the town of
Newport, Tennessee, and empties into the French Broad River at an elevation of 1,005 feet. The
slope of the river bed changes dramatically along its 80-mile path with the steepest portion in the
mountains near the river's source. The land cover surrounding the river varies from heavily
wooded to residential and agricultural (EA, 1987).
The flow of the river is altered at approximately river mile 38.1. Here Walters Dam was
erected which forms Waterville Lake(also know as Walters Lake). This reservoir extends for
several miles above the dam. The water is taken from the lake and runs through a tunnel to river
mile 26 where a powerhouse is located. After the water is used for hydroelectric power generation
by Carolina Power&Light (CP&L), it is released just prior to the Tennessee border (EA, 1987;
CP&L, 1994). The volume of water released at the powerhouse at any given time is dependent on
the volume of water entering the lake and the electrical demand placed on CP&L.
For accurate water type classification, the Pigeon River has been divided into several
segments. The portion of the river that flows through the mountains above the town of Canton
intake is classified as a WS III Coldwater trout fishery. This rating indicates that the water body is
a water supply for drinking, culinary or food-processing purposes. WS III's are protected supplies
with low to moderately developed watersheds. Beginning at the town of Canton (river mile 63) the
river is classified as a class C warm water fishery. This designation specifies use for secondary
(non-contact) recreation, aquatic life,wildlife and agriculture. This portion of the river is not
4

CD
td
e 03460766 TCt111t SSLC co �.
North Carolina
CD
� n3•Ir•4s0o w
• 03459780 rf.
o U3
a4� w 03453500 w <'�+ F-•
43464000
�ro
D3457124
UI A v} �°r• I��q7
A 03451926 �4X aa.
4 w vA JJJ" 03461976 ,�..
03416991 47 • 034515DO f+ w w �1,pl1
Y`f
03455500 03451000 � w 03463021 �{�.QO� Q1� 034633W
A
�
A 03447/140 03463160
03443000
AO Ambient Monit011ug StOtion
03439000
fnY
wwlue.+w a...ur.

Wallen Dam
ARM
Cataloocheehee
Creek Walters Lako
OHM 42.6
Creek
New Hepco Bridge, ff in
es
`\ d�RM 42.7
`fLOIN ARM 45.1�
ORM 48.2 „yy
ARM 48.3
A-,, C
ARM 46.(1. Ferguson Bridge f~D
Crabtree p
Jnn.rllians Creek
Creek
ARM 49.8
ARM 52.2
ORM 52.3
Clyde STP
ARM.... ARM 57.1 Fiberville A RM 62.9
0 AM 63.0
OHM 69.0
a
Waynesville M Mill Oulfall�
STP OR g ARM 63.3 G� Cantors, NC
54.
ARM 68.4 Richbnd
Creek m OHM 64.6
Clyd.:
❑RM655
Plots Farm
Addition
EA Engineering, Science and TechnOlOgy

French Broed River _
~A RM 2.3
V Newporl STP
ARM 4.0 C
'7
Newport, TN
OHM 7.6 b
Corby Geck
Willow Springs
ARM 13.6
i OHM 19.3
TENNESSEE
fLOW Hutlord
RM 24.9 % NORTH
'� CAROLINA
ARM 24.7 ARM 26.1

classified for primary recreation or drinking water(EA, 1987). In order to maintain these ratings,
water quality must be sustained at current levels so as to neither hinder, nor prohibit designated
uses.
The hydrology,topology and water quality of the Pigeon River all play important roles in
determining the magnitude of effects caused by Canton mill discharges and aid in understanding the
conditions surrounding these effects. At Canton, the mean monthly Pigeon River flow ranges from
174 to 310 mgd (270 to 480 cfs), except during September when the historical mean monthly flow
is 60.1 mgd (93 cfs) (EA, 1987). On average, since the Canton Modernization Project (CMP), the
mill uses and discharges less than 29 mgd (44 cfs) of water(Figure 1I1). The Pigeon River is the
sole source of mill water and the only discharge basin for the mill. Especially during the summer
months, mill effluent can be a substantial portion of the river flow. Therefore, to meet the stringent
limitations of the NPDES permit. strict process control and waste treatment is employed.
8

Figure III
SEFLOW (mgd)
O O O O O O O
Jan-87
Apr-87
Jul-87
Oct-87 —
Jan-88
C
Apr-88 m
Jul-88 �
Oct-88 �
Jan-89 -
Apr-89
Jul-89 m
Oct-89 z z m M
3 m
Jan-90 �,a CM 3
O v 9
Apr-90
Jul-90 ` ..
r r z
Oct-90 COc m
Jan-91 Ao 0
e �c c
Apr-91 5 Pe 3 n
Jul-91
Oct-91
S
Jan-92 T e
Apr-92 m
Jul-92
Oct-92
'o
Jan-93 t 9
Apr-93
Jul-93
Oct-93 } a
Jan-94
Apr-94
Jul-94 4-
Oct-94 =,
4-
Koelsch. Susanne. Effluent Discharge Graphs. Champion International Corporation. 1994.

III. Sources of Color
According to standard effluent guideline limitations proposed by the Environmental
Protection Agency (EPA), the paper industry can be broken down into twelve different mill types.
These types are based on the method of pulp processing and the products produced at each mill
(EPA Fact Sheet, 1993). The Canton mill falls into the Bleached Papergrade Kraft and Soda Mill
Subcategory which:
"Includes production of a bleached kraft wood pulp using an alkaline
sodium hydroxide and sodium sulfide cooking liquor. Principal products
include papergrade kraft market pulp, paperboard, coarse papers, tissue
papers, uncoated free sheet, and fine papers, which include business,
writing and printing papers" (EPA Fact Sheet, 1993).
The Canton mill has its own pulping facilities and four paper machines which produce a
variety of products. Three of these paper machines, numbers 11, 12 and 20. produce uncoated free
sheet printing papers, business papers, and envelope papers. The fourth machine, the number 19
machine, produces bleached paperboard (a material used in the manufacture of milk cartons and
juice cartons) (Champion Fact Sheet, 1994).
Color can be produced in a variety of ways within a bleached kraft mill. It is generally
believed that the color from the kraft process is generated as a natural by-product from the
aromatic portions of the lignin molecule during pulp production (Ziobro, 1990). According to
National Council of the Paper Industry for Air and Stream Improvement, Inc. INCAS]) research,
"Spent pulping and bleaching liquors contain varying amounts of
dissolved extractives, carbohydrates, and lignin. Although all three
components may contribute to color, lignin and/or its degradation
products are probably responsible for most of the color in pulping and
bleaching effluents" (NCASI #239, 1970)
10

At this point in time,the problem of relating color to a specific chromophoric group is still under
discussion. Prior to projects such as the Canton Modernization Project, theories believed that both
pulping and bleaching liquor color resulted from aromatic and quinonoid nuclei and carbonyl and
ethylene group structures. Structures such as these exist alone or as part of larger molecules
(NCASI #239, 1970).
These chromophoric resonant structures have been found to be sensitive to pH alterations,
easily oxidized or reduced and possibly agglomerated or metal-complexed. All of these alterations
can modify the intensity of the color produced by the chromophores (Lang). Recently. new
theories have proposed that a majority of the color from the pulping process is actually a product
of carbohydrate degradation which occurs during the cooking of sugars (monosaccharides and
hemicelluloses) in the kraft process (Ziobro, 1990). Regardless of source, the kraft color(known
as brown color or black liquor color) is known to have a high molecular weight (Koelsch, 1994;
Ziobro, 1990). The weight of the black liquor is due to lignin extracted from wood in the pulping
process. Current research indicates that the size and weight of these color bodies may play a key
role in any attempts to identify the fate of black liquor color. Because these black liquor color
molecules are so large it is highly unlikely that they can be taken inside the biosolid cell wall when
color is adsorbed.
As mentioned above, in addition to the kraft color, color is produced at other stages of the
pulp and paperrnaking process. In particular. color is produced during the bleaching stage. This
color is referred to as bleach plant color. Since the CMP, bleach plant color appears to have
significantly different properties than the brown color. According to past NCASI research, the
color of bleaching effluent comes from one or a combination of the three following sources:
1) chromophores which are dissolved unchanged.
2) some or all of the original chromophore groups which arc destroyed.
11

3) new(and possibly different)chromophores that are created (NCASI #242, 1970).
Before the CMP was completed, the majority of color from the mill came from the caustic
extraction stage of the bleaching process (NCASI 4239, 1970; Stratton, 1986). Today. the
majority of color entering the wastewater treatment plant is brown color from the pulp mill and
associated recovery processes since the technology changes made in the CMP have dramatically
reduced bleach plant color loads.
One other color producer in the mill was recently examined. This source is the dyes used
in the actual coloring of the paper. The Canton mill makes a variety of colored papers, each of
which requires a separate combination of dyes. According to research conducted during the
summer, these dyes would have no significant impacts on wastewater color, even in the case of the
worst possible spill scenario. For a better understanding of these various color sources, refer to
Figure IV, which contains a schematic diagram of the mill that shows the pulping, bleaching and
papermaking stages of the pulp and paper mill processes.
12

From . .
�pp Ca+rq Trr.rr
Chemical Additives
�- m-•� and Reactions
r .sr,..q ��.�j�,�1�1[+1== I'• -
,I•^,• _ na.r -�;,ti ___ �! I. • n Whila Ilquo,added to word
b,.nnl I w "1•:5:' r"ww r Uq.u.r I I (• clops- xlarts In dissolve I-grun
Irn lrwl kce,` 1 rlurlYry � '� (. .
so-.n,F rr...,R,r,l VVV'•'In�"�^�+ U� , np.ry.pe.enr -- M''`11`, Rb S!crwm
`R'� reb' . . =cl�.i.w-'•MJ u�.'.•.-wln�'a;i r n�T'I'•R!q-,.�¶Ir'rr(S(�
j
11 e liquor liqu
or I lignin
c residue)washed(while
r: Oxygen Hnlo,ludhnr
dlignificalion.T/
Il Pulp bleached to increase
-� � -•� F.Pn.w whiteness
It I nr W CMmk n.O .ry CTt•
I Pulp extracted with caustic soda
�.J I FlI E C b to further remove lignin.
•-It , I Pulp bleached In increase cr)
5 wholeness r
wl {� •Ir- '- I Bleached pulp slurry Pape,
P.... additives(such as PH adjusted
__ __
o.raaaim--I '"� ' -�' dyes.rosin,alum,clay,litannnn
if i+i H j�> l� / I �w � _ (' dioxide)added to enhance
'\•..Q �/ 1 I 1 �i I rI paper properties
•'+ ��� 1�' • �I • 11 Pape,sheet formed here
through dewalering
�•9 1� D n}.� , ull.r ra l I Black liquor concentrated
r
.w.wr I Black liquor converted to
green liquor-
'v 1 •w j /G%� FNwDM
�����` ✓%// n.. k Green liquor manuleclmerl
I Green liquor converted to
% while liquor.
1• v.'/
1.1 Calcium carbonate settled out
•e..1 and sent to lime kiln. Clarified
while liquor goes back into new
cooking cycle at Point A
N The temperature at non-conlacl
process wale,is reduced by
the tooting lowers for reuse in
the mill.
V Champion
Cb pvu nml be.lnahmral Corpmalion

IV. Historical Color Loads
The color loads from the mill to the Pigeon River have decreased dramatically over the
past several years due to the Canton Modernization Project. Table 1 shows that there has been
approximately a 77%decrease in secondary effluent color (secondary effluent is the wastewater
which has been treated and is discharged to the river) from 1987 to July, 1994. (See Appendix A.)
This same trend can be seen in Figure V which offers a comparison between key completion dates
in the CMP project and changes in color loading. A relatively small portion of the decrease in
color loading to the river is due to the decrease in over all effluent flow (mgd) to the river(Figure
III). Between 1992 and 1994 the secondary effluent flow from the mill has decreased from close to
45 mgd to between 25 and 30 mgd.
Table 1. Secondary Effluent Color
SmondSryEMueill SecondaryEttluent"
Year Color(scut Color(lb/dayl
1987 944 346,826
1988 11W4 379.972 _
1989 901 332,532
1990 798 300,757
1991 815 306,794
1992 681 243,004
1993 424 119,909
1/94-7/94 357 79,247
(Datastream, 1994).
14

Fieure V
SE True Color (lbs/day)
8
Jan-87 i
Apr-87
Jul-87 M
Oct-87
a
Jan-88
Apr-88
Jul-88 3
Oct-88 ..
A
Jan-89 0
Apr-89 n
Jul-89 r6
Oct-89 V.
Jan-90 wi 07
Apr-90CL
: 3
Jul-90 ' ' A B
Oct-90
Jan-91 F m^ M p a
r r. r 7
Apr-91 _t c e �- N ,
Jul-91 --
+ Nor_
Oct-91
Jan-92 : N
Apr-92 g
Jul-92
A
Oct-92 s
Jan-93 -�
Apr-93 3
Jul-93
Oct-93
A
Jan-94 ;
Apr-94 } ;
Kocisch, Susanne. Effluent Discharge Graphs. Champion international Corporation. 1994.
is

V. Significance of Color
When examining the effects of treated industrial wastewater on the environment, there are
a number of significant issues to consider besides color. However, color is a particularly
significant issue for two reasons:
1)aesthetics and community interest
2) potential impacts on ecosystem productivity.
Each of these issues is important and warrants further discussion and research. It appears
that with respect to color, the aesthetic value of the Pigeon River has been more closely examined.
To date, several cost benefit analyses have been conducted on the river. Each of these studies has
attempted to assess the visual, existence, and recreational values of the Pigeon River.
The Pigeon River is surrounded by several similar rivers and nearby communities which
are benefiting from the income from river recreation. According to some studies, if the aesthetics
of the Pigeon River were improved, recreation along the river could substantially increase. Some
studies indicate that profits from Pigeon River activities would rival those currently found on the
Hiwassee, Ocoee and Nantahala Rivers -- all key recreation sites (Bach and Barnett, 1987).
Debate is still occurring on the exact value of these monetary benefits. Cost benefit
analyses have assessed the financial benefits (from recreational use)that would occur due to
improved river conditions and the ability of Champion International Corporation to pay for these
improvements. The two most widely cited reports are "Benefits and Costs from the Reduction of
Color Effluent from the Champion Mill into the Pigeon River" prepared by National Economic
Research Associates, Inc. (NERA) and a combined document which includes "An Economic
Impact Analysis on the Recreational Benefits of a Restored Pigeon River" and "A Financial
Analysis of Champion International Corporation's Ability to Provide for a Clean Pigeon River" by
16

Orville Bach and William Barnett of Walters State Community College (NERA, 1988, Bach and
Barnett, 1987). These two reports rely on somewhat subjective criteria (such as graded variables)
and offer differing opinions regarding the financial benefits which would arise from improvements
in river color. Bach and Barnett indicate that with a "clean" Pigeon River the "total benefits over
next 10 years" are$18 million and the "total impact on economy" is $73 million (Bach and
Barnett, 1987). The NERA report states that the total benefits from reduced color discharges into
the Pigeon River fall between$10.9 and$17.6 million (1988 dollars) (NERA, 1988). At best, the
economic impact results from the two studies differ bya factor of four.
Upon examining the methods and findings included in the two studies, it is apparent that
both reports present highly subjective results. Bach and Barnett attempted to calculate monetary
benefits based on a Water Resources Council point system. The assessment graded environmental
quality, accessibility, recreational experience, and several other variables. Monetary values were
assigned to each of these factors depending on their point totals. The NERA report examined three
different color reduction scenarios for the river and also analyzed the Bach and Barnett study. The
NERA report refuted many of the point totals estimated in Bach and Barnett as well as many of the
initial assumptions in the report. For example, NERA states,
"Bach and Barnett assign a score of twelve out of eighteen for availability of
opportunity. A twelve would indicate no floating opportunities within one hour,
which seems to ignore the French Broad and Hiwassee Rivers. We find four to be
a more objective classification"(NERA, 1988).
Based on such statements and analysis of many other environmental impact statement
point/ranking methods (Nichols, 1988) it can be said that assigning points to determine recreational
values and deciding profits based on those values is, at best. an inexact science.
Despite differences, each study offers some interesting insights into the color issue.
However, with the passage of time, both studies are becoming less relevant to the current situation.
17

First, each of these studies was conducted prior to the Canton Modernization Project that made
major alterations to mill operations and significantly reduced effluent color loading (Figure V). So,
much of the color reduction which the reports speculate on has already occurred. Thus far, neither
team has returned to assess how the recent color load reductions have actually affected local
economies. Such a study would be extremely useful in determining the actual benefits from the
color reduction that has occurred. Second,a new cost benefit analysis conducted under the post
CMP conditions would be informative. Because, although the overall conclusions of these studies
are open to debate, they offer important insight as to the possible economic benefits that arise from
color reduction.
Just as benefits are difficult to predict, so, too are the costs of achieving those benefits. At
the time of the two studies, the Canton Modernization Project had not begun and the color standard
for the river had not been set. However. NERA analyzed several situations including the scenario
that eventually became the NPDES permit standard. This is a standard of 50 standard color
units(scu) at the Tennessee border(Yost, 1992). NERA estimated a cost to Champion of$113.3
million dollars to reach this color discharge level and benefits of$17.3 million from the reduction.
This is a significant difference in cost verses benefit. Now. $330 million later. the Canton mill is
meeting the 50 standard color units standard. But, the proportion of this expenditure which
actually went toward color reduction has never been identified. And. the question remains. "What
have been the economic benefits on the river from these expenditures?"
Without a doubt,major reductions in the effluent color discharged to the Pigeon River will
have aesthetic and recreational benefits. This is already being experienced since the CMP. During
the summer of 1994 there were several whitewater rafting companies operating on the Pigeon River
(Tennessee Department of Environment and Conservation. 1994: Bach, 1994). Because the Pigeon
has easy access off of Interstate 40, a major travel route, it may eventually compete with rafting
18

business on the Ocoee and Nantahala Rivers (Tennessee Department of Environment and
Conservation, 1994). However, questions remain as to what will be the cost to achieve any further
reductions. And, does the benefit cost ratio warrant such measures?
VA. Aesthetics
There are several reasons the aesthetics of the Pigeon River have been so carefully
examined. First, the mill is located on one of the smallest rivers utilized by a bleach kraft mill.
Because of the lower flow volume of the river, its dilution capacity is lower than that of other,
larger rivers. The color changes on the Pigeon River due to discharges from the Canton mill are
higher than the color changes in most other rivers throughout the country which are utilized for mill
discharge. But, this is not due to high color of mill effluent (NCASI, 1994). Rather, Champion's
Canton mill has one of the lowest bleach kraft mill effluent color loads according to the National
Council of the Paper Industry for Air and Stream Improvements, Inc. (Table 11).
Table II. Effluent Loads From Bleach Kraft Mills
Annual 7Q10 Co In(scu) Color(scu) Effluent Effluent
River River Location (r;7Q]0 (rrt�Avg Color(ppm) Flow
Flow(cfs) Flow(cfs) Company Town state River River 3/]/94 (MGD) River
Flow Flow Estimate
45 2.6 Georgia Pacific Palatka FL 1.341 788 1400 37 Rice Creek
77 8 P.H.Glatfelter Spring Grove PA 415 113 800 12 Codurous
Creek
233 67 Buckeye Florida Perry FL 1022 474 1900 50 Fenholloway
327 53 Champion* Canton NC 133 34 300 27 - Pi eon.R.
385 40 Willamette Johnsonburg PA 53 6 400 1 4 Clanon R.
395 47 Appleton Papers- Roaring Sprg PA 41 6 265 5 Juaniata R.
524 40 Boise Cascade• Deridder LA 226 20 1400 5 Sabine R.
710 70 westvacoo Luke MD 341 46 1150 19 Potomac R,
745 94 Westvaco' Covington VA 241 41 800 26 Jackson R.
896 190 Mead Escanaba MI 105 27 460 36 Escanaba R.
1086 200 S.D.Warren' Westbrook ME 89 18 649 21 Presumscot R.
1755 3.7 International Paper Bastrop LA 4279 105 4672 26 Bay.
LaFourchc
1766 27 Champion LuIlkin TS 687 20 1400 17 Angelina R.
2300 784 Champion Quinnesce MI 22 8 720 16 Menomnee R.
2353 301 Potlatch Cloquet MN 88 12 1400 13 St.l ouis R.
2676 539 Georgia Pacific- Woodland ME 83 18 1100 29 St.Croix R.
2853 208 Weyerhaeuser Nov Beni NC 93 8 640 23 Neuse R.
3226 730 Leaf River• New Augusta MS 17 4 400 19 Leaf R.
3555 668 Weyerhaeuser Oglethorpe GA 44 8 1683 12 Flint R.
3642 252 Container Corp of Amer lBrewton AL 203 16 1400 28 Conecuh R.
3732 1343 International Pa Jay ME 27 10 623 40 Androscoggin
19

3732 1343 Boise Conde' Rumford ME 42 16 1277 30 Androscoggin
3761 752 Mead Kingsport TN 1 9 2 430 10 Holston R.
4590 296 Mead Chilichothe OH 1 128 10 890 32 Scioto R
4620 691 Georgia Pacific Crossett AR 1 105 17 1400 36 Ouachta R.
4840 879 Bowater' Calhoun TN 1 76 15 1000 47 Hiwassee R
4939 1182 Nekoosa Port Edward WI 54 13 1400 30 Wisconsin R.
4939 1182 Nekoosa Nekoosa WI 54 13 1400 30 Wisconsin R.
4958 1363 Consolidated Papers WI Rapids Wl 23 7 1000 21 Winconsin R.
5029 1304 James River Berlin NH 20 5 953 18 Androscoggin
5269 863 Bowater' Catawba sC 67 11 1400 28 Catawba R.
5329 991 Union Camp Eastover SC 14 3 1000 9 Wateree R.
5644 634 Federal Paperboard Riegelwood NC 129 16 1400 42 Cape Fear R.
5997 151 Temple Inland Evadalc TX 494 19 1420 52 Neches R.
6549 3730 Stone Container Port Wenlwrth GA 7 4 1000 17 Savannah R.
7023 1027 Weyerhaeuser Pylmouth ME 91 14 1400 46 Roanoke R.
7087 1240 S.D.Warren Hinckley ME 46 9 1065 36 Kennebec R.
9810 1343 Willamette Bennetsville SC 42 6 862 45 Pee Dee R
10019 2424 Federal Paperboard AuRusta GA 20 5 1 1400 23 Savannah R.
11633 1 4007 ISimpson Paper Anderson CA 7 3 1400 14 Sacramento R
11732 234 lWeyerhaeum Columbus MS 85 2 942 15 Tombigbee R.
11965 3032 Lincoln P&P' Lincoln ME 7 2 1400 10 Penobscot R.
12195 2793 1 Pope&Talbot Halsey OR 11 3 1400 14 Willamette R.
13343 3596 JBoise Cascade Intemtl Falls MN 5 1 600 20 Rainy R.
13698 2124 JITT Ra oner Jesup GA 57 9 1400 58 Altamaha R.
14139 1786 JKimberly Clark Coosa Pines Al- 65 9 1400 57 Coosa R.
15138 3765 IJames River Old Town MR 8 2 1319 15 Penobscot R.
17184 780 lNekoosa Aslhdown AR 74 4 850 48 Red R.
23237 733 IGulf Stales Deena Iois Al. 45 1 1000 22 Tombigbee R.
26229 5232 lHarnmermill Selma AL 14 3 1648 28 Alabama R.
26231 1055 IBoise Cascade Jackson AL 31 1 1100 20 Tombigbee R.
29851 1495 James River Butler AL 63 3 1450 44 Tombigbee R.
32982 6549 Alabama River Claiborne AL 12 3 1000 55 Alabama R
41673 1141 International Paper Pine Bluff AR 72 2 1672 22 Arkansas R
50094 12927 Potlatch Lewiston ID 6 1 1400 33 Snake R.
51581 9159 Champion Counland LA 7 1 750 51 Tennesse R.
128496 9995 Willamette Hawesville KY 2 0 1400 11 Ohio R.
182384 79123 Boise Cascade Wallula WA 1 0 2000 18 Columbia R.
199256 44087 Westvam Wickliffe KY 1 0 1400 25 Mississippi R.
200073 89952 Janes River Clatskanie OR 0 0 640 37 Columbia R.
200073 89952 lBoise Cascade St.Ilelens OR 1 0 1000 35 Columbia R.
200073 89952 Weyerhaeuser lank view WA I 1 1400 49 Columbia R.
200073 89952 Longview Fibre Ianh vices WA 2 1 1400 65 Columbia R.
200073 89952 James River Cams WA 2 1 1778 51 Columbia R.
464386 109988 James River St.Francisville LA 0 0 588 27 Mississippi R.
464386 108988 Georgia Pacific ZLha'y LA 1 0 1400 27 Mississippi R.
556901 119132 Potlatch McGhee AR 0 0 1400 1 13 Mississippi R
587707 132724 International Paper INatchcz I MS 1 0 0 781 139 Mississippi R.
Total Mills=68.
'Effluent color is addressed in permit limits,monitoring,special studies,or other arrangements involving regulatory agencies
**Mill process water is highly colored and comes front below mill discharge
Note: Major estuarine in southeastern states,ocean intermittent discharge,and wine lake situations excluded.
(NCASI, 1994)
Several additional factors draw attention to the river color. The first factor is the location
of the Canton mill and the Pigeon River. Although other rivers throughout the country may have a
higher color than the Pigeon River, they are located where the aesthetic beauty and preservation of
rivers and ecosystems are not as omnipresent as they are amidst the national forest and park lands
20

of Western North Carolina. In many regions of the country, people are use to rivers and streams
with much higher background colors and numerous effluent loads. Surrounding Canton, natural
areas and relatively low stream color are the norm rather than the exception, and the environment is
subject to much public scrutiny.
As mentioned earlier, location is a key factor in another respect as well. From the Canton
mill to the Tennessee border is less than forty river miles. Because this is such a short distance, the
color effects from the mill are still measurable as the river winds its way across into Tennessee
border. Tennessee residents have expressed concerns in the past regarding the effects of the water
color on river recreation and productivity.
VR EcosystemProductivity
The next issue of importance is the potential effect of wastewater color on river
productivity. A number of tests have been conducted on the Pigeon River and nationwide to
determine if and when color loads significantly alter the biota and animal species found in a water
body. A relevant series of nationwide tests was conducted by the National Council of the Paper
Industry for Air and Stream Improvement. Inc. near New Bern, North Carolina from 1976 to 1981.
These studies examined the "Effects of Biologically Treated Bleached Kraft Mill Effluent on the
Periphyton Community in Southern Experimental Streams..." (NCASI #421, 1984; NCASI #423,
1984) Because the studies were conducted on a warm water stream in the Southern United States
(and the Pigeon River below Canton is classified as such), the conditions are more applicable to the
Pigeon River than are other studies that were either conducted on cold water streams or saline
environments. However, one must be careful in assuming that the results at the Canton site would
be identical to results at New Beni. The New Beni site is clearly a warm water stream located on
21

the Carolina coast. The Pigeon River originates as a mountain stream, which just above Canton is
classified as a cold water system.
In these studies, periphyton was used as a biological indicator because it is believed that...
"periphytic and planktonic algal communities may play an important role
in the oxygenation of streams, particularly in warmer summer months,
through the release of oxygen during photosynthesis. During the same
process, carbon dioxide is fixed as organic compounds which may enter
the foodweb and contribute to the productivity of the stream as a food
resource for macroinvertebrates and fish" (NCASI 4421, 1984).
During the first of these studies, conducted in 1976-1977, no difference was found in the
growth or production of fish populations when color up to 200 scu was added. Under 7Q 10 (the
ten year seven day low flow) conditions, Champion's effluent changes stream color by
approximately 260 scu at the mill effluent site and by 120 scu at Hepco. North Carolina.
approximately 21 river miles downstream (Koelsch, 1995). In addition. the study measured
periphyton concentrations at the depths of 30 and 90 centimeters and found that at these depths
neither the biomass nor the community structure of the periphyton changed. However, the study
did suggest that further research should measure the periphyton growth at points between these two
depths (NCASI 4421, 1984).
In another study, conducted at the same location between 1979 and 1981, the mean color
concentration was increased from approximately 70 scu to 260 scu of true color. This study found
that there was no clear correlation between effluent concentrations or color additions and the dry
weight of periphyton collected at four sample depths. Where the highest color additions occurred,
no difference was seen in chlorophyll a biomass between the control and experimental streams until
depths of 30 to 45 centimeters were reached. However, even the change in chlorophyll biomass
which was seen at these depths was not reflected in the total biomass of macroinvertebrate
secondary producers. The study concluded that "any differences noted in the overall production of
22

periphyton as measured by cholorphyll a biomass did not appear to affect the total biomass of the
macroinvertebrate secondary producers..." (NCASI 9423, 1984).
EA Engineering, Science and Technology, Inc. conducted a study in 1987, titled "Synoptic
Survey of Physical and Biological Conditions of the Pigeon River in the Vicinity of Champion
International's Canton Mill." The study assessed the actual changes in biological conditions along
the Pigeon River due to the color discharges from the Canton mill (EA. 1987). This studv
investigated then current(pre-Canton Modernization Project) biological and water quality
conditions in the Pigeon River and modeled potential changes which could occur if Canton mill
effluent color were reduced. EA found that the operation of the hydroelectric power plant at the
Tennessee border had a significant influence on the patterns of available light downstream of the
border. It appeared that the structure of the benthic community downstream of the power plant
was actually due to the flow fluctuation caused by operation of the plant (EA. 1987). The
modeling showed that reductions in color of 50 and 90 percent of the existing level could increase
the amount of light reaching the bottom in both riffles and pools. In the vast majority of pools (and
most riffles), light conditions would improve with a 50 percent reduction in color. In contrast, with
a 90 percent reduction in color,the bottom light intensities would inhibit optimum algal growth in
most areas (EA, 1987).
The study also concluded:
1) An appropriate balance is occurring between trophic levels.
2) Chlorophyll a and biomass of periphyton communities both reach maximum
values where color levels were near the maximum measured in the river.
3) Benthic biomass increased and taxonomic variety decreased downstream of the mill
discharge (as compared to conditions upstream of the mill).
4) The Index of Biological Integrity(IBI) indicate that the mill has a definite effect on the
fish communities below the mill discharge. However, the lowest IBl ratings were not
correlated with the highest existing color. Therefore, there is no reason to conclude that it
is color which is responsible for the reduced IBI's (EA, 1987).
23

Since the time of the report, the color loading to the river from the Canton mill has
decreased by approximately 77 percent in terms of pounds per day of color. See Appendix A for
calculations. Therefore, if the predictions made in the EA light attenuation models are correct,
light conditions in the river should have improved significantly between 1987 and 1994. As
mentioned, the EA study predicted a decrease in river productivity when color decreased by 90
percent. Therefore, further color load reductions may not improve stream periphyton productivity.
It is important to note that any changes that have occurred in the river productivity may be due in
part to other changes in mill operation which transpired concurrently with color discharge
reductions. Color is one of many variables that changed during the Canton Modernization Project.
VI. Governine Policies and Permittin¢
To maintain stream quality. there are policies which govern both dischargers into the river
and stream conditions themselves. The Pigeon River water body classifications which assess and
regulate stream conditions were discussed earlier. Although the Canton wastewater treatment plant
is the major discharger into the Pigeon River subbasin of the French Broad River Basin (Figure
VI), there are 61 other wastewater treatment facilities which also discharge into the Pigeon. There
are four major (> I mgd) dischargers on subbasin 05 of the French Broad River Basin. Thcv are
Dayco Corporation, Waynesville Wastewater Treatment Plant. Maggie Valley Wastewater
Treatment Plant, and Champion (Clark, 1993). Figure VI shows the locations of these facilities.
Each of these dischargers, including Champion, is governed by a National Pollution Discharge
Elimination System permit. However, even when all the dischargers are considered, Champion and
24

Major NPDES dischargers(>1 MGD) •„.
• AB other NPDES dischargers ;;.,„ ;•;•,
111 I {3r11111111; Il
'11^ ••.III { -�
�\y
�yi�r�.. �, `yam\` r' S't wY•' .IIIYI11111
<Vi' / 1„,C � .%•J.. I 4•a IP n'�rr� 11.1'_
" ''Y a•ir •j , '\ IIFe\\11 •� �.Z� � ,lr, N
Or
�,i^/II l // i' f 1- l n, l• - ,y.., a ,�
�1C 1'l,\ fl it 'I• Jr,.l ,al IIII1'
T i Jy 4 4' •4 i •C L
4l• �".... ril
\a, .Nnl 11 1 " t� 1., i . .. " • i _ r nuuml I r n Ic.
ullnni, {•' � 1 •. J;.l l'. e J nmll
1 C� Ir ❑ , •.I Jr.l.• �' ,L4 ,J J," •'% '1\I Y 1 '
_ i �•� _�Illll /aL' `•tr L. 4 " �` C<`• 1 � .1 Jy
1 it N .:rilv 4 ;d®rn M1 'i' ` a/1-.lY� fi J11 \-�_�,((([,Ilu ' ia, •'J'
Olr Ylll Yn,lnl ,ili _I F' ` tl• Spl•
uuunrl ull v{•'ll ru+(V,� a.-� -
�/�� pOiq a/ .'t`" \� ,7 1ii q �i •��`.. �!'_
l •
< \\{\': •, rL~t ' ^ t t'+a • 's illy
�i �, „ ,,;;• • r • r..,,"�!: � r• L. Key to Major NPDES Discharges
" ..'v+ •'i ` ••I' "✓'� }�y i �•..,.., ::'�'.+--\ Mm g D15C116[ger Name Mlle M Discharger Nme
y o.i"Iup { �- '`!, , ,'^" t' nonrn•n n,\• l Drevmd WWTP 9 Champion
2 Ecusta 10 Dayco Industries
iri.. 1 t li Lr r \ u rN {� 3 DuPont I Waynesville WWTP
4 General Eimtrk 12 Maggie Valley W W7T
C, ,,,,,`ySF n.11. a l •••n. 1 lendersonville
'•Tt� Jill,i �C%I 6 Cranston Printwarlu 14 K-T Pelds ar
3P
�' � '� 7 CP&L- Asheville 13 Unimlo
• auNs `"
* . '4 9 BASP
br
Division of Environmental Management
Distribution Map of NPDES Dischargers in the French Broad River Basin

the Waynesville Wastewater Treatment Plant still make up the majority (greater than 80%) of the
wastewater flow to the Pigeon River(Clark. 1993). Although current discharge estimates for all
the facilities are unavailable, it is estimated that Champion makes-up approximately 70 to 85
percent of the wastewater flow to the Pigeon.
During the summer of 1994 when these color studies were conducted, mill effluent was
controlled by a NPDES permit that became effective on October 24, 1989 and was effective for 5
years. The permit was issued and later revised by the United States Environmental Protection
Agency (EPA). Originally, the permit allowed for a discharge flowrate of 48.5 mgd. Since the
completion of the CMP, the limit has been revised. The current discharge allowance is 29 mgd
(U.S. EPA, 1989).
The color standard was a main point of contention in the Canton mill's NPDES permit.
Over the years the EPA has changed its position on the issue of federal color standards for the pulp
and paper industry. In 1974 the EPA issued a technology based color standard, but in 1982 this
standard was revoked. The EPA declared that color was not an issue of national importance and
should be regulated as necessary on a case by case basis (Yost, 1992).
Most recently, a new set of regulations was proposed for the pulp and paper industry.
These regulations are based on a 1985 lawsuit filed by the Environmental Defense Fund and the
National Wildlife Federation against the EPA. This suit claimed that the EPA was not effectively
regulating dioxin and furan discharges from bleach kraft pulp and paper mills. This case was
decided in favor of Environmental Defense Fund and National Wildlife Federation. The court
stated that not only did the EPA have to regulate dioxin and f iran discharges, but the agency also
had to assess the "multiple pathway risk"due to sludge. wastewater and other pulp mill products.
So, the new regulations (known as the "Cluster Rules") control both air and water pollution
26

produced by the pulp and paper industry. One of the many proposed standards in the Cluster Rule
is a monthly average color discharge limit of 76.3 pounds of color/1000 pounds of pulp produced.
The Canton mill is already operating well below this proposed limit (Figure VII).
The North Carolina standard, under section 15A NCAC 2B .0211(b)3(F) of the
administrative code, states that point source color cannot interfere with the designated uses of a
river. The color shall not harm public health, aquatic life or wildlife, secondary recreation, the
palatability of fish, aesthetic quality or the water for any designated uses (Clark, 1993). The color
standard which prevails in this case is for recreation. The regulation states. "There shall be no
turbidity or color in such amounts or character that will result in any objectionable appearance to
the water" (Yost, 1992).
Since it is impossible to define "objectionable appearance" and regulate under this
standard, the law requires:
"Where the state's only water quality standard for a particular pollutant is
articulated in a narrative fashion, the permit writer is obliged to translate
such standards into a numerical limit on a case-by-case basis" (U.S. EPA,
1992).
It is on this numerical standard that debate occurs.
In the case of the Pigeon River, the determination of an adequate color standard was a"
fairly intricate procedure that spanned several years. The color standard developed for the Pigeon
was a result of a collaborative effort by the state of North Carolina, the EPA, and the state of
Tennessee. The first step was the determination of background color levels in the river. Color
measurements upstream from the mill were found to average 13 scu, so this was considered
background color. Next, the state of Tennessee was asked by the EPA to design a model
discharge permit for Champion that specified by how much the Canton mill should be able to alter
27

Fi¢ure VII
SE True Color Obs/1000 Ibs pulp)
O s ui a :n
b b o b
Jan-87
Apr•87 y >
Jul-87 _ 7
Oct-87 +
_ •e
G
Jan-88 n �
Apr-88
R
Jul-88 ..
n
Oct-88 3
Jan-89
Apr-89 ,
Jul-89 no n
Oct-89 S •: QQ c
A S.
Jan-90 e'
Apr-90 e_
Jul-90
Oct 90 ° e C6 e
Jan-91 a e
A
Apr-91
p 9
Jul-91 m
Oct-91 _ A �_
m
Jan-92 C6
Apr-92 '
Jul-92 3
Oct-92
e
Jan-93
Apr-93 '®
Jul-93 n
Oct-93 Q
Jan-94
Apr-94
Jul-94'
Oct 94
28

river color. The Tennessee reply recommended that:
"In-stream apparent color not be increased above background by more
than 40 platinum-cobalt color units when measured outside a mixing zone
extending from the discharge to river mile 48.2 at Ferguson Bridge I i.e.,
below the Tennessee-North Carolina state linel" (Yost. 1992).
Shortly thereafter(May 14,1985) the state of North Carolina issued Champion's discharge
permit. However,the question still remained as to what percent removal should be required of
Champion. Following the objection by Tennessee to the color limit in the North Carolina permit,
the EPA overruled the permit and assumed control of the permitting process. In September of
1989, after several years of litigation and public hearings. the EPA issued the final operating
permit for the Canton mill. This permit has a color standard of 50 scu at the state line on a
monthly modeled basis (Yost, 1992) This standard was later upheld in court after being challenged
by both Champion and a citizens' group known as the Dead Pigeon River Council (Yost, 1992).
Twice a week color measurements are obtained from grab samples taken at the state line.
Therefore, a model is used to calculate state line color more frequently. This model provides a
daily comparison of state-line color to the standard, as required by the permit.
The North Carolina/Tennessee state line color sfandard(SLc) calculation/measurement is
based on an equation included in the NPDES permit for the Canton mill. This Equation is:
SLc= (WTPe/8.34) + ((HEf-WTPf)*Dc)
Hef*10 E ((-0.224 * Log HEf) + 0.781)
Where:
SLc= Instream true color at North CarolinafTennessee border(state line)
WTPc= Monthly average waste treatment plant discharge color, calculated as the average
of all daily loading values (expressed as pounds of true color per day) for a calendar month
HEt= Monthly average flow (mgd) at Hepco, North Carolina -daily values lower than
81.4 mgd shall be entered as 81.4 nrgd
WTPf= Monthly average waste treatment plant discharge flow (mgd)
29

Dc= Color concentration of all Dilution streams (13scu)
(Champion, 1994)
This equation can be compared to actual measurements over the last several years to assess its
accuracy. Thus far, the equation results and actual test values have been comparable.
As mentioned above, there are a number of policies (and proposed policies) that are or may
effectively assess and govern the color of the Pigeon River. These are:
1) The Canton Mill National Pollutant Discharge Elimination System
Operating Permit
2) Department of Environmental Management Water Quality
Standards
3) French Broad River Basinwide Plan
4) EPA "Proposed Effluent Limitation Guidelines and National
Emission Standards for Hazardous Air Pollutants for the
Production of Pulp, Paper,and Paperboard" (Cluster Rule)
Although the NPDES permit currently is the most stringent control on the Canton mill
color discharges, this may not always be the case. The evaluation and regulation of the Pigeon
River are in a state of flux. Currently, three of the four forms of evaluation and regulation for the
Pigeon River water quality are being updated. The Canton mill National Pollution Discharge
Elimination System permit is up for renewal. The French Broad River Basinwide Plan is currently
under revision. And, the Cluster Rules have not been made final and effective. Therefore, all the
regulations and methods of evaluation are important in governing river conditions. Should river
conditions or regulations ever change significantly, the water quality standards. the French Broad
River Basinwide Plan or the Cluster Rules could provide important information and result in new
standards for controlling mill discharge. In addition, other forms of regulation, such as the Clean
Water Act Amendments, could also play a role in regulation.
With all this in mind it is important to consider what information is necessary to assist the
regulatory process. Afterall, it is important to regulate a river based on current information.
30

Before any further changes are made in mill or river regulations it may be prudent to assess current
mill and river conditions.
VII. Wastewater Treatment
The wastewater treatment facility at Champion's Canton mill treats all the effluent from
the mill before discharging it into the Pigeon River. In addition to the mill wastewater flow, which
averages approximately 25 -29 million gallons per day, the facility also treats the town of
Canton's municipal wastewater. (The municipal wastewater is pretreated with chlorine before
entering the wastewater treatment facility.) On average. the flow from the town is approximately
0.6 million gallons per day (less than 3 percent of the total treatment plant discharge).
The wastewater treatment facility consists of a series of basins and mechanical, biological
and chemical treatment mechanisms. (See Figure VIII.) As wastewater enters the facility, it
passes through a grit chamber and bar screens which remove larger objects that could harm the
plant pumping system. From there the wastewater moves, with the help of the low lift pumps. on
through the three primary clarifiers. After settling occurs in the primary clarifiers. the wastewater
flows into one of the four aeration basins. Ammonium hydroxide, N144OH, and phosphoric acid.
H3PO4,are added as nutrients between the primary clarifiers and the aeration basins to aid in the
treatment process. In addition, sulfuric acid, H3SO4, may be added for pH control, and calcium
carbonate, CaCO3. is added intermittently. In the aeration basins the wastewater was combined
with polymer and aerated. (In October. 1994. polymer use was discontinued.) From the aeration
basin, the water passes into the secondary clarifiers. Here final settling occurs and the biosolids
are returned to the aeration basins. Last, the flow passes through a flume and over a reaeration
cascade before being discharged to the river.
31

During treatment, the color removal across the treatment plant is approximately 30
percent. The research presented here is a laboratory analysis of the factors affecting color removal
across the Canton wastewater treatment plant. A variety of factors have been considered in the
analysis, and several have proved to have a significant affect on color removal. Based on these
results, and the results of related works, several suggestions have been made for further areas of
study.
32

Sampling Sites:
PI-Primary Influent
PE-Primary Effluent
Wastewater Treatment Plant Flow Diagram SE-Secondary Effluent
., AB-Aeration Basin
` Bypass Waste Mixed Ll uor
o H2SO4 pH Control Nutrient Feed I
'w (Back-up) I r -nB or - - - - - '
N OH H3PO4 DlAester Aeration
n AB
2.3 G r 3.4 MG
r
Primary Effluent r r r
AB or r Aeration r Polymer Addition
02 k0cllonkitemYller> suftconto I Dji�estery Basin r For TSSControl
I 1- - 2.9MG 3.4MG r w
Intermittent AdditionPE - - - - - - - - - j
Clarifier
#if I of CaCO3 Aeration Barn LL
I- - -
200' Diameter - - ' - - - - - 1 - - - - - - _ _ _ 20 Surface Aerators /4
14,75' Deep - - '
r Clarifier
3.47 Million d
Dewatered 200' Diameter
Gallons Sludge 12'Deep
r 2.82 Million
` r Gallons
%
` Primary Sludge
- - - - - - - - - - - - - r Recycle Sludge r
Four - - - - - - - - - - -
r Arm-Andrih r .
/2 l Belt Presses r `
Clarifier r Flow l `Q
200 ' Diameter Splitting r #5 m
14.75' Deep r Station I r Clarifier
3.47 Million 200' Diameter L SE
Gallons r r 12' Deep o
r Press r 2.82 Million a
r Filtrate r Gallons r
r °u
i l n
Clarifer
125' Dlamete % #6
1.01 MG Clarifier
11'peep CO2 pH 150'Diameter
Control 1.65 MG
14'Deep I
1.01 Million Gallons Oxygen
Luofiw PI a C Grit 1.85 Million Gallons 2w
Pumps m m Chamber LL
'X Reaeratlon
City Seale Cascade N
MIII Wastewater City
Fr. FLUME
Final Effluent for Flow Compllance

EXPERIMENTAL RESULTS

VIII. Obiectives of Research
Because a high percentage of river flow is provided by the mill treated effluent, color
removal at the Canton wastewater treatment plant(WTP) is closely monitored. On a day-to-day
basis, changes in mill and wastewater treatment plant operations can (and do) cause the percent
color removal occurring across the wastewater treatment plant to vary. To gain a better
understanding of the effects of these operational changes the Environmental, Occupational Health
and Safety Department at the Canton mill set out to study the color removal processes at the
wastewater treatment plant. Several research objectives were outlined prior to the initiation of the
study. These goals were:
1) to gain a better understanding of the current color removal processes at the wastewater
treatment plant. (Are the processes physical or chemical'? What type of color is the
wastewater treatment plant removal affecting? What factors are significant in the color
removal process?),
2) to determine the fate of color(Is it becoming part of the biosolids either physically or
chemically?),
3) to evaluate if physical color removal accounts for a significant percentage of the total
color removal across the wastewater treatment plant.
4) to analyze the role of metals in color removal. and
5) to determine if the phosphate buffer added during pH adjustment interferes with the
measurement of color.
IX. Areas of Study
To achieve these objectives. studies were conducted over a twelve-week test period that
began May 16. 1994. and ended August 5. 1994. The majority of research took place in the
laboratory facilities of the wastewater treatment plant at the Canton mill. However, some of the
metals analysis, which required more advanced technology, was conducted at the corporate
34

technology headquarters in West Nyack. New York. Samples were shipped up to West Nvack
several times a week for analysis.
Three types of experiments were tun over the course of the summer. The first of these
experiments were the phosphate buffer tests. There has been some concern in the pulp and paper
industry that the phosphate buffer used in the pH adjustment of wastewater color test samples
interacts with calcium in the samples and interferes with true color measurements. Initial concern
regarding this issue was raised in 1986 by David Bonistall of Champion International (Carpenter,
1986). National Council of the Paper Industry for Air and Stream Improvement, Inc. investigated
these concerns and found that when wastewater containing high levels of soluble calcium was
combined with the phosphate buffer, precipitation occurred and color test results were significantly
lowered. NCASI researchers suggested that because of this phenomenon the phosphate buffer
should not be used when high levels of calcium are present in the wastewater (Carpenter. 1986).
However, the exact amount of calcium necessary to cause color test interference was still
somewhat unclear. To determine if this type of interference is occurring at the calcium levels
normally present at the Canton wastewater treatment plant, Champion decided to conduct its own
phosphate buffer tests. These tests were run primarily to assure the validity of later test results.
The second type of experiments were desorption studies. These tests were used to
determine if a portion of the color removed by biosolids at the wastewater treatment plant could be
desorbed back off the solids when the concentration gradient of the wastewater solution was
altered, a strong metal chelating agent was added, or a strong basic solution was added. These
tests were conducted based on the hypothesis that much of the color removal which occurs across
the wastewater treatment plant is due to the physical adsorption of color from wastewater onto the
biosolids. If the experiments showed that physically desorbing color off the biosolids was possible.
then the hypothesis would be validated.
35

The final set of experiments was a variation of the desorption experiments. In these
reactors color was initially desorbed off of the biosolids by changing the concentration gradient(as
done in the desorption experiments). Then, the concentration gradient was again altered by adding
one of a variety of color sources to the reactors. This second step was an attempt to readsorb color
onto the biosolids. Finally,the gradient was reversed once more to desorb color off of the
biosolids. These experiments involved a three step process and will be referred to as the desorb-
adsorb-desorb experiments.
X. Techniques of Color Measurement
To achieve uniformity in color measurements for all wastewater samples (including those
for these experiments), a set procedure has been designed for testing true color. Samples are
compared to platinum-cobalt standard solutions with known color concentrations to achieve color
measurements. It should be noted that the above procedure is for the measurement of true color.
Wastewater samples can also be tested for apparent color. Apparent color is the color measured
in an unfiltered sample. Therefore, apparent color measurements are highly dependent on size,
type, and make-up of the particles in the sample. No apparent color tests were used in these
experiments, and the standards which control color discharges for the mill are in terms of true
color. In preparation for the test. samples are adjusted to a pH of 7.6 using hydrochloric acid
(HCI) and/or Sodium Hydroxide(NaOH). This is done due to the fact that color measurements are
known to be highly pH dependent. At higher pH levels color test readings can increase
significantly.
A phosphate buffer is added to assist in the pH adjustment of color samples. This
buffering has created some controversy. As mentioned, several industry studies (Carpenter. 1986)
have indicated that the phosphate buffer could possibly interact with calcium contained in the
36

samples and interfere with test results. This issue was addressed in several of our experiments and
will be discussed later in this paper.
After pH adjustment, the sample is filtered through a 0.8 micron porosity membrane filter
in order to remove turbidity. If the sample is not properly filtered, sample turbidity can
overwhelmingly influence the tests, cause false high color readings and make the test results
invalid. When turbidity is especially high (such as directly following a storm event) samples may
need to be centrifuged in order to accurately measure true color.
Next, samples are ready to be tested on a spectrophotometer, a machine that measures light
transmittance through a sample. As mentioned above, the machine is standardized either using
platinum-cobalt standards of 100, 500 and 1000 scu (standard color units) or 10,100 and 500 scu.
The spectrophotometer which is standardized with the 10, 100 and 500 scu samples is used for
lower range color samples and gives more precise color readings. This machine was utilized for
the majority of color samples run during the summer experiments. Once the spectrophotometer has
been standardized, samples are run and light transmittance is measured. True color measurements
are obtained using a conversion from light transmittance to color(NCASI, 1971).
X1. Buffer Experiments
The phosphate buffer experiments will only be discussed briefly at this time. The data
from these tests relate to other experimental results and will be mentioned again later.
The buffer tests were conducted over a five-day period. Each day a 24-hour composite
sample was taken of the primary influent, primary effluent, and secondary effluent at the
wastewater treatment plant. (See Figure IX for the exact locations where sampling occurred.) The
wastewater samples from each of these locations were tested with and without the phosphate
buffer. Each sample was pH adjusted and filtered according to NCASI color test specifications.
37

As can be seen in Table III there appeared to be no visible trends in color measurements due to the
effects of the buffer. The samples with buffer showed no pattern of having consistently higher (or
lower) color measurements than the samples which did not contain the phosphate buffer. The
difference in measured color between the buffered and non-buffered samples appears to be
explainable by random color test variation. It was concluded that at normal Canton mill
wastewater calcium levels color precipitation did not occur, and the phosphate buffer does not
appear to interfere with color test results. (According to these studies, normal aeration basin total
calcium ranged from 110 mg/I to 170 mg/I and soluble calcium was approximately 54 mg/I.)
Table III. Phosphate Buffer Effects on Color Test Results
Primary Primary Secondary
Influent Effluent Effluent
Color(scu) Color(scu) Color(scu
W/o w/o w/o
Da w/butler butler w,rps w/bull'er huller I W'1'1" wlbutler I butler Wfl'
1 755 755 493 722 733 537 601 583 i06
2 537 528 5112 5fi4 564 502 537 546 451
3 443 443 435 451 459 583 411 379 387
4 470 485 510 519 528 5U2 379 379 395
5 485 493 519 451 468 451 357 379 383
'"MY'column contains treatment plant operators daily color measurements. Note that first day's readings do not
agree with WTP operator readings. This is probably due to miscalibmtion of the spectrophotometer.
XII. Desorption Tests
XII A. Desien and Methods
The next type of tests were a series of bench-scale reactor experiments, which measured
the ability to desorb color off biosolids. These experiments varied slightly in design, but the
general make-up of the experimental reactors was similar. The experiments were broken down into
three basic types:
I) desorption reactors that used only tap water to desorb color from the biosolids.
38

2) desorption reactors which utilized a strong chelating agent for metals, EDTA (tetra
sodium ethylene diamine tetra acidic acid), to desorb color from biosolids, and
3) desorption reactors which utilized an alkaline solution of sodium hvdroxide (NaOH) as
an agent of color desorption.
The experimental design for these tests was loosely based on work conducted in 1986 by Steve
Stratton of Champion International at the Canton wastewater treatment plant (Stratton, 1986).
Before beginning this round of experiments, research was conducted into possible
explanations for the color removal mechanism. It appeared that much of the color removal that
occurs across the wastewater treatment plant is a result of physical adsorption of the color bodies
to the biosolids (Stratton, 4/9/86). Adsorption is most often a result of two factors. First,
adsorption occurs when a solute is lyophobic. In other words, the solute(in this case the color
bodies) is not attracted to the solvent. The less water soluble a substance is, the greater the
accumulation by the biosolids (Tsezos, 1986). Second, the solute has a strong attraction to the
solid (Weber, 1972). The rate of adsorption is based on many variables including temperature.
pH, particle diameter, concentration of solute and molecular weight of solute. Generally. the
higher the molecular weight of the solute the easier it is adsorbed. Color producing bodies, as
mentioned earlier, are known to have a high molecular weight (Eckenfelder, 1966).
There are two types of adsorption which occur: physical adsorption and chemical
adsorption. Adsorption reactions betnvecn cell walls and metals can occur due to electrostatic
attraction, covalent bonding, ion pairs. complexes and chelating. But, electrostatic attraction is the
most important of these forces. The biosolid cell walls and the polysaccharides that are attached
have a strong anionic charge. This charge attracts the positively charged metals such as calcium
and magnesium. Usually, multivalent cations will be bound preferentially over monovalent cations
with low atomic weight. But, calcium and magnesium are known to compete with other metals for
sorption sites (Henderson. 1994). When physical adsorption has occurred. the adsorbed particle is
39

usually not attached to one specific site on the solid, and the particle is free to move around on the
solid surface. With chemical adsorption, the molecules are not free to move along the surface of
the solid (Eckenfelder, 1966). According to Eckenfelder, ofien adsorption of organic molecules
occurs due to "specific interactions between identifiable structural elements of the sorbate and the
sorbent." As adsorption occurs, particles will continue to be adsorbed onto the solid until dynamic
equilibrium is reached. The surface area of the cell walls is an important factor in determining this
equilibrium (Henderson, 1994). At this time the rate of adsorption will roughly equal the rate of
desorption and the concentration of solute remaining in solution will reach an equilibrium.
Based on this information and previous experiments, it would appear that a significant
portion of the color molecules are physically- adsorbed onto the biosolids. This adsorption most
likely occurs with the assistance of calcium (or other multivalent) ions. Past tests would indicate
that the color molecules bind to calcium ions (Stratton, 1986). Then these calcium ions provide the
key necessary to allow the color molecules to attach onto the biosolids and be removed from the
wastewater solution.
The current experiments were designed to test this hypothesis and determine if physical
adsorption is still a predominant mechanism of color removal given the recent operational changes
in the mill during the Canton Modernization Project. If color could be desorbed off the biosolids
using only tap water, not only would it show that color had been physically adsorbed onto the
biosolids, but it would also indicate that these color molecules had undergone no chemical changes
in the process.
In the second type of desorption experiments, EDTA was used to chelate the calcium in
solution. If the calcium was indeed important in the adsorption of color, then chelating calcium
would prevent the adsorption of color onto the biosolids and pull adsorbed calcium and color back
into solution, thereby increasing the color of the solution.
40

Finally,NaOH was used to test the effects of pH changes on color adsorption. As
mentioned, adsorption is highly pH dependent (Eckenfelder, 1966). Higher rates of adsorption
occur at lower pH's. The NaOH raised the pH of the solution significantly. This increase in pH
was expected to increase desorption.
In preparation for these tests, a single 24-hour composite sample (taken at 15 minute
intervals by a programmed sampler) of mixed liquor suspended solids (MLSS)was obtained from
the aeration basins of the wastewater treatment plant for each experiment. MLSS is a term
referring to the combination of wastewater and biosolids contained in the aeration basins. (See
Figure Vlll for aeration basin (AB) sampling location.) Each of these samples was allowed to
stand for approximately 30 minutes to one hour to allow the biosolids portion of the mixture to
settle. The settling time for each sample varied according to the sludge volume index of the mixed
liquor suspended solids. The liquid portion of the mixture ( which was termed "simulated
secondary effluent"or "SE") was siphoned from the top of the settled sample and used for control
reactors. The remaining portion of the sample was a concentrated mixture of biosolids and
wastewater and was termed "simulated return activated sludge" or --RAS". Experimental reactors
were set up containing one liter of RAS and 2.5 liters of tap water. (The initial few rounds of
reactors were designed to contain RAS to water ratios which varied from 1:2.5 to 1:30. Although
it is important to have a high enough ratio of water to RAS so that desorption can occur, it was
found that the color change in the higher ratio reactors was so small that it was impossible to
measure. Therefore, the ratio of 1:2.5 was used in all later reactors.) EDTA or NaOH was
substituted for a portion of the water in reactors which contained these chemicals.
The control reactors, which contained SE with little or no biosolids,were also tested.
These reactors held 1 liter of SE in place of the I liter of RAS. The controls were set-up to show if
any color changes were occurring which could not be attributed to the presence of biosolids.
41

The desorption reactors were aerated in plastic vessels for a period of up to 96 hours.
Color measurements were taken at several different stages. First, the color was measured in the
undiluted aeration basin 24-hour composite sample. A theoretical color could be calculated based
on the dilution ratio in the reactors. (In a 1:2.i reactor. original reactor color= original sample
color/3.5.) Next, the color was measured in each of the reactors initially. (Color measurement
usually occurred about one hour after dilution. This delay was due to set-up time and test
preparation time.) In many of the reactors, especially those containing EDTA or NaOH, the initial
color readings were significantly higher than the theoretically calculated color. Since this was not
the case in the control reactors, this difference is attributable to the fact that desorption can be a
fairly rapid process (Amy, 1988). It is quite likely that color desorption occurred during the time
between reactor set-up and initial color measurement. After the initial measurement, color
measurements were taken approximately every 24 hours for the duration of aeration. The reactors
were run for periods ranging from 24 hours to 96 hours. The aeration time simply depended on
what day of the week the reactors were set up. The final color reading for each reactor was taken
on Friday. After the final reading, the reactors were emptied and cleaned in preparation for the
following week's experiments.
XH B. Results and Discussion
The results of the desorption experiments are summarized below in Tables IV. V. and Vl.
Table IV presents the results from the reactors that used only water to desorb color from the
biosolids. A significant color increase was seen in each of the experimental reactors. (See the
column labeled "% Change".) The control reactors, which contained no biosolids, actually showed
a slight decrease in color measurements over time. (See Appendix B for the control reactor
42

results.) These results would indicate that it is possible to physically desorb color from the
biosolids using no chemical mechanisms.
Table IV. Summary of Color Chan es in Water Reactors
Actual Color Actual Color
Ratio 'Meorelical (scu) (scu)
!<Swnp1eDate RAS:Water Colo scu Initial Final %Chan e
5/22-5/23 1:5 85.0 111.0 1 131.9 55%
5/22-5/23 1:10 46.4 65.0 79.5 71%
5/22-5/23 1:15 31.9 43.0 54.7 71%
5/22-5/23 1:20 24.3 50.0 35.2 45%
5/22-5/23 1:30 16.5 22.0 23.9 45%
5/24-5/25 1:2.5 95.4 104.9 171.2 79%
5/24-5/25 1:5.0 55.7 62.1 99.1 78%
5/24-5/25 0.5 39.3 44.3 676 72-A
5/24-5/25 1:10 30.4 3M98
1 85%.
6/5-6/6 1:5 55.9 6 50%
6/6-6/7 1:5 40.4 ( 841A
6/7-618 1:5 45.4 4 42%
6/12-6/13 1:2.5 78.9 7U 76%
6/13-6/14 1:2.5 71.8 8 60%
6/14-6115 1:2.5 69.4 7 56%
6/21-6/22 1:2.5 88.1 11 .3 51%
6/24-6/27 1:2.5 94.4 110.2 127.4 51%
6/27-6/28 1:2.5 76.2 90.0 120.6 58%
6/28-6/29 1:2.5 79.7 88.5 124.0 56%
7/2-7/5 1:2.5 87.7 93.3 125.7 43%
7/5-7/6 1:2.5 87.5114.8 31'%
7/10-7/11 1:2.5 85.5 90.2 III.9 3PA
7/IU-7/11 1:2.5 85.5 81.7 IIG.9 37%
The results of the EDTA experiments can he seen below in Table V. The volume of
EDTA used in the reactors varied. The most commonly used volumes were .02, .05, 0.1 and 0.2
liters of.05M EDTA per 3.5 liter reactor. (However, due to supply problems. during the first
week .02M EDTA was used.) The reactors were designed such that if. for example. .05 L of
EDTA was used. then the reactor would contain I L of RAS. .05 L of EDTA and 2.45 L of water
Further experiments could be conducted to detennine the EDTA saturation point at which greater
concentrations of EDTA would not increase color changes further. Throughout the EDTA
experiments the reactors experienced increases in color from 165 to 500 percent. These changes
43

are significantly greater than the changes in the reactors containing no EDTA (See Table IV).
Because the color changes in these reactors were large, questions arose regarding the sources of
this color. It was suggested that the EDTA might somehow be reacting with either the biosolids or
the liquid, thereby creating chromophoric compounds. This possibility will be discussed in greater
detail later.
Table V. Summary of Color Chan es in EDTA Reactors
.:: 'Actu51
%Color(scu)
Satgple Vol! Theorehcel-:':
pate .. :HDTA(ml) Color(ppm) .-q Hours —18 Hours —48 Hours —72 Hours %Change
5/31-6/1 10' 151 164 163 8%
5131.6/1 50• 151 1 44 772 14a/,
5/31-6/1 100• 151 217 199 32%
5/31-6/1 2500 151 260 290 92%
6/5-6/6 10 168 330 445 165%
615-6/6 50 168 348 739 340%
6/5-6/6 100 168 356 623 271%
6/7-6/8 10 136 362 574 322%
6/7-6/8 50 136 348 622 357%
6n-6/8 100 136 339 466 2430A
6/12-6/13 20 138 317 595 331%
6/12-6/13 100 138 304 655 375'%
G/12-6/13 200 138 356 543 293%
6/14-6/15 20 121 377 688 469%
6/14-6/15 100 121 429 734 507%
6/14-6/15 200 121 427 700 479%
6/24-6/27 20 148 170 558 277%
6/24-6/27 100 148 317 461 2II°/a
6/24fi/27 200 148 415 459 20A
•In the experiments dated 5/31.6/1.02M EDTA was used rather that.05M EDTA.which was used in all other experiments.
The final type of reactors. which are shown in Table VI. contained either 10 or 20
milliliters of NaOH. These reactors showed color increases similar to those measured in the
EDTA reactors. Color increases were on the order of 100 to 325 percent. This increase would
appear to validate the theory that pH increases cause greater desorption to occur. To further
understand the processes occurring in the reactors, several reactors were desorbed with water for
44

several days before the NaOH was added. Through this type of experiment. it was possible to
observe what additional color,which had not been affected by the water desorption, was being
extracted from the biosolids by the NaOH.
Table VI. Summary of Color Chanties in NaOH Reactors
Actual Color
Sample Theoretical (scu)
Date . ,.::;Goloc:(scu). O;Hrs 4-5 Hrs .8:Hrs I8 Hrs 48.1irs 5313rs 96 Hrs Change
5/31-6/1 151 240 520 244%
6/12-6/13 N/A 139 296 113%
6/12-6/13 N/A 139 326 135%
6/13-6/14 N/A 115 334 190%
6/13fi/14 N/A Hs 395 243%
6/21-6/22 N/A 133 277 108%
6/21-6/22 N/A 133 283 113°/�
6/21-6/22 88 101 298 239%
6/21-6/22 88 108 353 301%
6/24-6/27 N/A 127 250 97%
7/5-7/6 88 103 224 155%
7/10-7/11 86 94 363 322%
N/A indicates that these reactors were desorbed with water prior to the addition of NaOH,so theoretical color is not
an applicable measurement.
Because of the huge increases in color seen in the EDTA and NaOH reactors, questions
arose regarding whether these color increases were really due to physical desorption of color.
Could some chemical reaction be occurring with the EDTA. for example. which was creating
chromophoric color compounds'? One possibility for the change was suggested. When the NaOH
or EDTA was added to the reactors it caused the pH to change from approximately 7 or 8 to 1 1. It
is probable that this dramatic increase in alkalinity caused the biosolid cells to lyse and spill their
contents. To test this theory, a physical method of lysing the cells was attempted to assure that no
chemical change was simultaneously taking place. The method used ecas ultrasound. In these tests
the original color of the RAS sample was measured. Then, the sample was subjected to varying
doses of ultrasound. Six to seven minutes of ultrasound proved sufficient to cause maximum color
change. In some samples, the RAS was diluted with water at a ratio of 1 unit RAS to 5 units water
45

before being subjected to ultrasound. To ensure that no chemical change would occur due to
temperature change, the temperatures of all samples were held below 100 degrees Fahrenheit. (This
is approximately wastewater treatment plant aeration basin temperature.) In order to validate that
cell lysing was indeed taking place, the July 10 samples were observed on the microscopic level. At
this level of magnification, it was evident whether or not the microorganisms were still viable. As
can be seen in Table VII, the color changes that occurred due to ultrasonic cell Ming are
remarkably similar to those which occurred due to the EDTA and NaOH addition. However,
questions arose regarding the validity of the ultrasound results. Because the cells had been lysed ,
many minute solid particles were present. Concern was expressed that these particles could be
passing through the .8 micron filter creating solids interference in the spectrophotometer readings.
In order to ensure the validity of the results from these experiments. the ultrasound samples tested
on August 4, were centrifuged and filtered in order to remove solids.
Table VII, Summary of Color Chanties in Ultrasound Experiments
Final Actual
Original Color Test Color Dilution Color
:Sam '1eDale Sample Descri tion (sell) (scu Ratio (scu) %Change
7/10 RAS-3 minutes of ultrasound 299 1 166 1:1 1 166 290°6�
7/10 RAS-6 minutes of ultrasound 299 387 1:5 1935 547°6�
7/12 RAS-6 minutes of ultrasound 349 1819 1:1 1919 421%
8/4 RAS-5 minutes of ultrasound 435 312 1:5 1560 259%
8/4 RAS-7 minutes of ultrasound 435 372 1:5 1860 328%
8/4 RAS 9 minutes of ultrasound 435 372 1:5 1186(1 328%
The final step of the desorption experiments was to attempt to quantitatively determine
what percent of the color removal from the wastewater was accounted for by physical adsorption
to biosolids. Unfortunately. these calculations produced wideh• varying results because of the
different residence time of the wastewater and the biosolids in the wastewater treatment plant
system. The calculations depended on the primary influent color, secondary effluent color, and
46

several other variables. Because the wastewater has an average total system detention
time(primary clarifiers, aeration basins and secondary clarifiers) in the system of 24 hours(with
three basins operating as aeration basins) and the biosolids have a seven day mean cell residence
time (MCRT) it was impossible to determine what averaging time should be used for the primary
influent and secondary effluent color loads. Based on the variety of averaging times used for these
calculations, the percent removal accounted for by these physical desorption experiments ranged
from 33%to 500%.
XI7 C. Conclusions
Several conclusions can be made from these desorption experiments. First. based on the
water desorption experiments, it can be concluded that a significant portion of the color removed
throughout the wastewater treatment plant is physically adsorbed to the biosolids without under
going chemical change. Second,given this fact, if the color load to the wastewater treatment plant
were to decrease dramatically over a relatively short time period, then it is quite possible that color
could actually desorb off the biosolids into the low-color wastewater. Therefore, the color decrease
that was seen in the primary influent may not be apparent in the secondary effluent color.
However, if the low-color load continued over a several-day (or longer) time period, the wastewater
treatment plant would reach a new color removal dynamic equilibrium.
It is possible to assume that the large increase in color that was seen in the EDTA or
NaOH reactors is explainable in physical terms. In the EDTA reactors, the increase in color was
most likely due to the reversal of the color concentration gradient. a decreased adsorption of color
as a result of the EDTA chelating calcium and a lysing of the cells due to increased pH. if calcmnt
actually binds the color to the biosolids (as it appears to). when the EDTA locks up the calcium
color desorbs off the biosolids but cannot be readsorbed. In the NaOH reactor the increase in color
47

was probably due to a reversal of the color concentration gradient and the lysing of the biosolid
cells caused by excessive alkalinity. However, in some cases, a significant pH change is known to
affect adsorption.
The EDTA, NaOH and ultrasound experiments appear to indicate that color is released
when the biosolid cell wall in ruptured. However, as mentioned earlier, the color molecules are
fairly large and therefore would be unable to enter the biosolid cell (Koclsch. 1994). Based on
these two facts, it must be theorized that the color bodies are tightly bound to the cell wall and are
only removed when the cell wall is lysed.
XIII. Desorb-Adsorb-Desorb Experiments
XIII A. Methods and Design
The final set of experiments were the desorb-adsorb-desorb experiments. During the
twelve-week research period three desorb-adsorb-desorb experiments were run. Each experiment
was designed based on the results of the previous sets of reactors and each lasted one week.
Several weeks after the three experiments were completed one additional set of experiments was
conducted. This last round will be included in the discussion and conclusions because the data
provide validation of earlier results and allow for some additional conclusions.
The desorb-adsorb-desorb experiments were an effort to gather more specific infomation
regarding color adsorption after a significant number of the desorption reactors had been run.
These studies were designed to provide several specific details. First, is it possible to reverse the
color adsorption process several times simply by changing the concentration gradient? Second.
does color removal across the wastewater treatment plant favor one color source over another? As
mentioned earlier, color comes from a variety of sources in a bleach kraft pulp and paper mill. The
48

two primary color sources of concern are the pulping stage brown color (represented in these
experiments by weak black liquor (WBL)) and the bleach plant color (BPC). Third, do metals play
a significant role in color removal? And, finally, what happens if the concentration of the metals in
solution are altered? The effects of metals on color were analyzed both with and without the
presence of biosolids.
The first round of reactors was set-up during the week of July 10, 1994. As mentioned,
these reactors were designed as a three-step experiment. First. the reactors were run identically to
the reactors in the desorption experiments. RAS was combined with water (at a ratio of 1:2.5). and
the reactors were aerated for 24 hours to desorb color from the biosolids. At the end of this time,
the color was measured and the reactors were allowed to settle. Then, the liquid was siphoned off
the top of the reactors, and a concentrated solution of biosolids remained. A color solution was
added and an attempt was made to readsorb color onto the biosolids by simulating current aeration
basin conditions. On the day the 24-hour composite sample was taken the RAS decant color was
measured at 298 scu. To replicate these conditions, the reactor target color was set at 300 scu.
A series of concentration calculations were used to determine how much color must be
added to each reactor. The volumes and colors of the biosolids remaining in the reactors were
known. The final volume was a constant at 3.5 liters and a final target color was set. (In this case.
the target color was 300 scu) A calculated volume of a known color source (either bleach plan[
color, black liquor color, or printary effluent) was diluted with a calculated volume of water (and
in some cases a colorless calcium solution). This mixture was combined with the biosolids to
achieve a 3.5 liter reactor at the target color. In a few cases large additions of the color source
were required. To achieve the target color, the reactors sometimes had a mixing volume of greater
than 3.5 liters. If this was the case. some of the solution was extracted so all reactors had a final
49

volume of 3.5 liters. A sample spreadsheet used for these calculations can be seen below in Table
Vlll.
Table Vill. Sample Spreadsheet for Calculation of Volume of Color Source to Add
Color Vol
Ugmd Btosolids Biosoiids. .Color Liquid 1Vater Vol Final
Color �.Reimining' Color Target Added Added Extracted Vol
Reactor Liquid seu liters (sell) (scu) liters (liters) (liters) (liters)
`.
1 PE' 357 1.25 111 300 2.55 tl 0.30 3.50
2 BPC•' 1216 1.25 108 300 0.75 1.50 0 3.50
3 wBL{a1 1182 1.25 116 300 0.77 1.48 0 3.50
4 PE+Ca0
357 1.443 too 300 254 0 0.48 3.50
°PE=Primary Effluent
••BPC=Bleach Plain Color
@WBL=Weak Black Liquor
NPE+Ca=Primary Effluent+Calcium
During the first week of desorb-adsorb-desorb experiments, four experimental reactors
were run. The reactor contents were:
R-1: 'RAS + Primary Effluent
R-2: RAS + Bleach Plant Color
R-3: RAS + Weak Black Liquor
R4: RAS + Primary Effluent + Calcium (276 mg/1)'
'276 mg/] of calcium was chosen for this experiment based on preliminary metals work that indicated that
the calcium concentrations in RAS were approximately 276 mg/l.
These combinations were chosen in order to observe if wastewater treatment plant color removal
favored one color source above another. If the color removal varied greatly between reactors R-2
and R-3 this would indicate that the removal process preferred either weak black liquor or bleach
plant color. Color change in the reactors which contained primary effluent would indicate if both
types of color, weak black liquor and bleach plant color(and/or other constituents in the mill
wastewater), were necessary for color removal to occur. Significant color change in R4, which
contained RAS, primary effluent, and calcium: would show if the addition of multivalent cations
could increase color removal at the waste treatment plant.
50

Calcium was used in these experiments to observe the affects of multivalent cations on
color removal. There are several reasons why calcium was used instead of other multivalent
cations. First, past experiments have indicated that calcium is a key element in the color removal
process occurring at the wastewater treatment plant (Stratton, 1986). Second, before these
experiments were begun we had already received some preliminary chemical analysis that West
Nyack had conducted for the desorption experiment reactors. As seen in Table IX, the numbers
indicated that calcium was found in the wastewater at much greater concentrations than were the
other multivalent ions. Additional metals results from the West Nyack analysis can be found in
Appendix C.
Table IX. Metals Analysis
?Al Ca il. Fe :K" M -': '.'Mn: SiZn " TQC�s.
RAS 6/27 Total 10.8 135 3.62 24.3 7.32 1.73 258 I3.6 5.57 0.52 NT
RAS 6/27 Soluble 20.1 190 5.37 22.4 14.1 3.24 500 17.4 8.65 1.02 130
SE 6/27 .48 37.1 0.37 17.8 5.03 0.73 382 1.5 2.98 0.07 45.7
'All concentrations are given in mg/l.
" NT indicates measurements which were not taken.
Normally,total metals content would be higher than soluble metals content. However,here the soluble
concentrations are larger than total. This anomaly is due to the filtering out of solids.
XIII B. Desorb-Adsorb-Desorb Experiment #1 Results and Discussion
The data from the first experiments did not provide any clear-cut results but did lead to
several observations that aided in the design of later, more conclusive experiments. As can be seen
in Table X, the only reactors that showed a decrease in color were those containing primary
effluent. In addition, the reactor that contained primary effluent plus calcium showed a 18 percent
greater decrease in color than reactor R-1, which contained primary effluent but no calcium.
51

Table X. Adsorption Results for Desorb-Adsorb-Desorb Experiment #1
AcluzillColor , Actiial..W(W*� Acfiial.Coiow
Theoretical 0 Hours 24 Hours 48 Hours
Reactor Sorb-Conditions Color(scu) (scu) (Scu) (scu) %Change
I PE 300 257 262 263 -12%
2 BPC 300 275 309 341 14%
3 WBL* 163* 200 224 223 37%
4 PE+Ca(2761ng/1) 300 250 218 125(210) -30%
Control Decant+PE+Ca 300 207(223)@ 168(218)@ 131(209)@ -30%
. (276mgfl)!#
•Miscalculations were made on reactor 3 which caused the theoretical color to be lower.'
1 No biosolids present.
# Possibly missed target color.
@ Numbers in()represent tests run w/o buffer.
The reactors that had additional calcium (R.4 and Control) had much lower color readings
when the phosphate buffer was added for pH adjustment. It appeared that color precipitation was
occurring when buffer was added to the samples containing additional calcium. To account for
this problem,the pH adjustment of the samples with additional calcium was run with and without
buffer. The result of the color tests without buffer are believed to be a more accurate
representation of true color and therefore are used in the calculations of-% Change".
After the final color measurements were taken for the adsorption portion of the
experiments, the reactors were allowed to settle. Then, the liquid portion of the reactors was
siphoned off. The remaining concentrated biosolid solution was combined with water to a final
volume of 3.5 liters Oust as in the desorption experiments). The initial reactor color was taken.
After a 24-hour aeration period, the final reactor color was measured. The results of the final
desorption stages in all the reactors for desorb-adsorb-desorb experiments 41, 2, and 3 can be seen
below in Table XI. (No final desorption experiment was conducted for desorb-adsorb-desorb
experiment 44.) In each of the three desorb-adsorb-desorb experiments, it proved possible to
desorb color from the biosolids in this final stage.
52

Table XI. Results of Final Desorption Staves for Desorb-Adsorb-Desorb
Experiments#1, 2, and 3
Color(scu) Color(scu) C61or(scu) Color(scu)
'"ReactorAdsoi•,b 77heoretical 0 lirs 0 Hrs 24 Hrs 24 Hrs
3Conditions Colorscu wBuf wtoBuP wBuf w/oBuf* %.Chan e!'
PE 118 117 131 11%
BPC 144 142 161 12%
WBL 71 83 78 10%
PE+Caf2761 72 68 68 81 76 6%
PE 37.4 37.1 74 98%
PE+Caj 1561 21.1 30.7 33.1 52.7 51.1 142%
Pe+Ca[2761 20.6 26.8 30 42.2 54A 164%
BPC 57.1 57.3 77.3 35%
BPC+Ca[1561 54.3 50.8 52.4 67.4 75.7 39%
WBL 40.7 45.1 59.3 46%
WBL 48.2 49.2 60.9 26%
WBL+Cal 1561 10.7 21.2 20.4 39.8 39.8 272%
WBL+Cal60l 17 26 49 188%
WBL+Cal 1201 12 24 45 275%
WBL+Ca[901 19 32 70 258%
WBL+Ca[901 20 34 65 225%
WBL+Ca[601 31 48 95 206%
VIBL+Ca[1201 27 42 109 304%
BPC+Ca 90 55 66 93 69%
BPC+Ca 90 54 GS 90 66.7%
' Reactors with high levels of soluble calcium were tested with and without buffer.
" "%Change"was calculated as(24 hour color-theoretical color)/theoretical color. For reactors that were tested
with and without buffer,the non-bullered final color was used to calculate"%Change'.
XIII C. Conclusions for Desorb-Adsorb-Desorb Experiment #1
Several conclusions could be made given the results from the first week of experiments.
The most important result appeared to be that color was not being adsorbed when the bleach plant
color and weak black liquor color alone were added to the reactors. However, adsorption did occur
in the reactor which contained primary effluent. This would lead to the idea that adsorption is
dependent on a combination of constituents coming from the mill. Next, we learned that calcium
increased the rate of color adsorption in the reactor which contained primary effluent. This would
validate the importance of calcium in the color adsorption process. Finally, in the control reactor,
it appeared that color was being adsorbed even when biosolids were not present. The observation
53

would seem to indicate that some interaction was occurring between the simulated secondary
effluent and the calcium which decreased measured color in the reactor. However, upon further
study it would seem that the lower color measurement may actually be due to ])missing the target
color when setting up the reactor (since the first color measurement was 223 scu and the following
two readings were 218 and 209) and 2) precipitation of color with calcium and the phosphate
buffer.
Xlll D. Desorb-Adsorb-Desorb Experiment #2 Methods and Design
The next round of reactors was designed based on the observations from desorb-adsorb-
desorb experiment#1. During the week of July 18, eight experimental reactors and four control
reactors were aerated. As before, during the first 24 hours color was desorbed from the RAS using
tap water. Then, reactors were drained of their liquid portion and the color sources were added.
The reactors contained:
Reactor Contents
I Primary Effluent
2 Primary Effluent + Calcium [156]
3 Primary Effluent+ Calcium [276]
4 Bleach Plant Color
5 Bleach Plant Color+ Calcium [156]
6 Weak Black Liquor
7 Weak Black Liquor
8 Weak Black Liquor+ Calcium 1 156]
In this experiment two levels of calcium were tested. These levels were chosen based on the
calcium levels found in initial MLSS samples (156 mg/1) and RAS samples (276 mg/1).
The four control reactors (C-9 ,C-10, C-11, C-12) all contained decant from the
desorption reactors in place of the biosolids contained in the experimental reactors. The control
reactor contents were:
54

Reactor Contents
I Decant + Primary Effluent+ Calcium [1561
2 Decant+Primary Effluent
3 Decant+ Bleach Plant Color
4 Decant+Weak Black Liquor
The results from these experiments can be seen Table XII. Only the reactors that
contained calcium showed color removal. The experimental reactors in which color adsorption
occurred were: those containing primary effluent, primary effluent and calcium or weak black
liquor and calcium. (Primary effluent, since it is a combination of all the mills wastewater, already
has.some-calciumin_it from-varioussources.throughout-the mill.) The color removal in the reactor
containing bleach plant color+ Calcium was negligible, and actually a color increase was seen
when buffer was added prior to color measurement. Of the control reactors, only the reactor
containing primary effluent and calcium showed a decrease in color, and this was probably due to
buffer interference. In the reactors that contained bleach plant color or weak black liquor(brown
color) but no calcium,the color actually increased. The greatest color removal was seen in the
reactors containing primary effluent + 276mg/I of calcium and weak black liquor+156 mg/1 of
calcium. The reactor containing primary effluent + 276 mg/I of calcium showed a 48 percent
decrease in color and the reactor containing weak black liquor+ 156 mg/I of calcium showed a 68
percent decrease in color. The greatest color increase (31 percent) was seen in the reactor which
contained decant+weak black liquor.
55

Table X11. Absorption Results for Desorb-Adsorb-Desorb Experiment 42
Ttieoreucal 0 Flours 24-14 ri 48 Hduri:.�
scul,
.-TWU :§6rbltondi dris- *,*,!'.�*�C'oIor.*'-*"'. �*;� � . �'( ( *u)...
1 PE 250 225 201 206 -18%
2 PE+.Ca11561 250 148 164 123(194) 40%
3 PE+Ca[2761 250 141 158 116(130) -48%
4 BPC 250 240 273 296 18%
5 BPC+Ca1156] 250 245 248 253(230) -8%
6 WBL 250 260 307 285 14%
7 WBL 250 252 296 293 17%
8 WBL+Ca11561 250 56 62 60(81) -68%
C-9 Decant+PE+Cal 1561 250 202 153 184 -26%
C-10 Decant+PE 250 236 257 268 7%
C-1 I De + IC 250 36 57 3%
C-1 t+ _25 27
1 317 32
7
31%
Results in Q are tests w/o the addition ot'jillospliale hulTer.
A target color ol'250 was used in au ell'ort to duplicate Ilieu current MLSS color conditions
XIIIE. Conclusions from Desorb-Adsorb-Desorb Experiment #2
An important result from these experiments was that the color removal by biosolids seems
to have had a much greater effect on the brown color than it had on the bleach plant color. The
change in color in the bleach plant color reactors is insignificant. It can also be concluded that
color removal only acts on brown color when calcium is present. Therefore, if weak black liquor
color entered the wastewater treatment plant and no calcium was present, then no color removal
would occur.
XHI F. Desorb-Adsorb-Desorb Experiment #3-Methods and Desken
A third set of desorb-adsorb-desorb reactors was designed based on the observations from
desorb-adsorb-desorb experiment#2 and the conclusions from other experiments conducted over
the summer. Once again, color was desorbed from the biosolids using tap water. After desorption,
56

the reactors were decanted and only the biosolids remained. The adsorption phase of the
experiment was based on a two factorial design with a centerpoint. The variables were color
concentration and calcium concentration. (See Figure IX.) The color concentration was varied in
order to ascertain if color removal changed as color concentration in the reactors increased or
decreased. The calcium concentration was varied in an effort to locate the breakpoint at which
calcium is the limiting factor in the color removal process. The objective was to estimate the level
of soluble calcium at which color removal ceased to occur.
_ _Figure IX
Reactor Design for Desorb-Adsorb-Desorb Experiment #3
500 • •
400 •
E
n 300
`0 200
0
U 100
0
50 70 90 110 130
Calcium Concentration (mg/1)
Figure IX shows the values of color and calcium tested in the experiments. The color was
tested at target concentrations of 250. 375 and 5110 scu. These concentrations were chosen because
they represent the typical color ranges seen in the full scale wastewater treatment plant. The
calcium concentrations tested were 60, 90 and 120 ntg/l. These concentrations were used because
they represent a spectrum of calcium concentrations below 156 mg/l and data returning from West
Nyack was indicating that average calcium concentrations in RAS and SE were below 156 mg/l.
From previous experiments it was known that color removal occurred at calcium levels at and
above 156 mg/l. Therefore, the limiting concentration of calcium must be below this point.
Since the previous experiments had shown that the color adsorption affected the weak
black liquor color and not the bleach plant color, the experimental reactors for the third week of
57

experiments only contained weak black liquor as a color source. (Reactors were run containing
bleach plant color and biosolids, but they were considered control reactors.) One reactor was
aerated at each of the color and calcium combinations shown in Figure IX, and the centerpoint
concentrations were run in duplicate. Each of the four control reactors (C-7, C-8, C-9, C-10) was
run at the experimental design centerpoint (375 scu of color, 90 mg/I of calcium). Control
reactors, C-7 and C-8, were duplicates of each other and contained bleach plant color, calcium,
water and biosolids. One reactor, C-9, contained bleach plant color, calcium, water and decant.
The last reactor, C-10, contained weak black liquor, water, calcium and decant.
XIII G. Results and Discussion - Desorh-Adsorb-Desorb Experiment #3
In each of the experimental reactors, color removal was approximately 60%. The few
percentage point differences in removal between the similar reactors is most likely attributable to
either color test error or differing aeration rates. For each of the experimental reactors, the color at
time=0 hours was well below the theoretical color. Since the dilution technique appears to have
closely approximated theoretical color in the past and in the control reactors. it is likely that color
adsorption was occurring very rapidly in the reactors.
The results from the control experiments were equally useful. In each of the control
reactors, the color increased slightly. The greatest color increase (26%)was seen in reactor C-9
which contained weak black liquor,water, calcium and decant. This result and the results from R-
6 and R-7 in Experiment 92 would indicate that color removal of the brown color does not occur
unless both calcium and biosolids are present. Similar results were seen in control reactors C-I I
and C-12 of the previous experiment. In the bleach plant color reactors, the color increased slightly
even though both calcium and biosolids were present. (See Table XIII.)
58

Table XIII. Adsorption Results for Desorb-Adsorb-Desorb Experiment #3
Actual i;'
Color
4 .,
-:Theoretical.:.
S 'b'C -d'i 4Reactoroi 0­ Chan e
I WBL+Ca[60] 250 195(204) 119(127) 101(101) _60%
2 WBL+Ca[1201 250 84(186) 85(100) 68(91) -64%
3 WBL+Ca[90] 375 220(314) 138(151) 127(137) -63%
4 WBL+Ca[90] 375 253(311) 138(160) 127(149) -60%
5 WBL+Ca[60] 500 455(472) 220(245) 210(220) -56%
6 WBL+Ca[120] 500 151(367) 169(193) 163(168) -67%
7 BPC+Ca[90] 375 388(390) 387(390) 395(392) 5%
8 BPC+Ca[901 375 390(404) 390(400) 380(398) 1%
9 WBL+Ca[90]+Decant 375 ___ 340(419) 398(406) 448(471) 26%
10 BPC+Caj90J+Decant 1 375 1 414(402) 1 416(432) 439(452) 17%
X111H. Conclusions from Desorb-Adsorb-Desorb Experiment #3
Because adsorption rates were approximately 60%we can conclude:
1) Color removal is significant in reactors containing biosolids, calcium and weak black
liquor color.
2) The calcium levels tested were still high enough for adsorption to occur.
3) No bleach plant color removal occurs, even in the presence of calcium and biosolids.
X1111. Desorb-Adsorb-Desorh Experiment #4-Methods and Desij?n
As mentioned, one final set of experiments was run at the Canton wastewater treatment
plant approximately one month after the other experiments were concluded. The experiments were
conducted by Susanne Koelsch and Bill Henderson of Champion. Information regarding these tests
was obtained from a report on the results written by Koelsch (Koelsch. 1994). This experiment
used variable concentrations of color and calcium as before. But. an additional variable. biosolids
concentration, was considered. The experiment was a two-cubed design with a centerpoint
59

replicated in triplicate. Several extra points were added to replicate some of the calcium
concentrations from the third experiment.
In all, seventeen reactors were run. The results appear below in Tables XIV and XV. The
biosolids concentrations were 1207, 2414 and 4827 mg/I of simulated return activated sludge. The
color concentrations were 250, 375 and 500 standard color units. The calcium levels were 30, 60
and 90 milligrams per liter. These calcium concentrations were chosen because the calcium
breakpoint was not reached at the higher calcium levels found in the last experiment.
Table XIV. Adsorption Results for Desorb-Adsorb-Desorb Experiment #4
(ma/1 of Color Removal)
:.Reactor-, . Color Target'::; TSS": :':Colormg/f } ' ;Color ->"color'- ' Color .` 'Color - '"Color Color
'Number'.:':Target CamgA mgQ : Theoretical 0- rem scu rem scu rem scu rem scu - rem scu remacu .f
scu RAS Hm OHM 24 Hm 48 Hrs O Hrs 24 Hm 48 Hm
w/o bur w/o bur w/o bur w/buf w/buf w/buf
1 500 90 4827 495 127 252 306 157 261 336
2 500 30 4927 495 -27 -58 63 -27 -31 97
3 500 90 2414 496 9 148 277 51 168 313
4 500 30 2414 496 -102 -237 -116 -90 -214 -95
5 500 90 1207 496 -84 -19 82 -70 5 139
6 500 30 1 1207 496 -168 -270 1 -200 .160 -260 -193
7 250 90 4827 247 87 128 150 109 138 164
8 250 30 4827 247 21 44 95 27 55 92
9 250 90 2414 247 77 120 134 92 133 157
10 250 30 2414 247 .13 -24 11 -6 -22 29
11 250 90 1207 248 13 91 123 28 127 150
12 250 30 1207 248 -56 -84 49 -50 -86 44
13 375 60 2414 372 -2 42 118 0 52 139
14 375 60 2414 372 18 so 119 22 59 139
15 375 60 2414 372 2 34 IUS -2 36 129
16 375 60 01
433 -174 •26R -22G -190 -24 -204
17 1 375 1 60 1 0 43.1 -1 G.1 -1RR -241 -1 R3 -195 -227
•TSS-Total suspended Solids
60

Table XV. Adsorption Results for Desorb-Adsorb-Desorb Experiment #4
( % Color Removal)
Color'. Tar&A. MS.? ."Golor.mpJl Color,. Color .Color 'CCo1or "':Color ' :iC61or
Reactor Target Ca mg/fF mrll.<, ::Theoretical •rem°(o. .. ...:rem% .. :ran /o .:re1n% rnn%
Number seu<. lw -O.Hrs 0 Hrs .'.:24 Hm 48 Hrs 6?1hs .;24 Hrs. 481:Hrs
f1.`.wlo buf: "iw 6uf a aw/o buf 'I:ivl&zf
1 500 90 4827 495 25.7% 50.9% 61.8% 31.8% 52.8% 67.9%
2 500 30 4827 495 -5.4% .11.6% 12.8% .5.4% -6.2% 17.6%
3 Soo 90 2414 496 1.8% 29.89/b 55.8% 10.3% 33.9% 63.1%
4 500 30 2414 496 -20.6% 47.8% -23.4% -18/2% 4312% .19.2%
5 Soo 90 1207 496 .16.8% .3.7% 16.6% 15.8% 1.1% 29.1%
6 500 30 1207 496 -33.8% -54.3% -40.2% -32.1% -52.3% -39.8%
7 250 90 4827 247 35.2% 51.8°! 60.7% 44.1% 55.9% 66.4%
8 250 30 4827 247 8.5% 17.8% 38.4% 10.9% 22.2% 37.2%
9 250 90 2414 247 31.3% 48.7% 54.3% 37.3% 53.990 63.6%
10 250 30 2414 247 .5.1% -9.6% 4.6% -2.3% .8.7% 11.9%
11 250 90 1207 248 5.2% 36.7% 49.6% 11.3% 51.2% 60.5%
12 250 30 1207 248 -22.6% -33.9% .19.8% -20.2% -34.7% -17.8%
13 375 60 2414 372 -0.6% 11.3% 31.7%. 0.0% 14.0% 37.4%
14 375 60 2414 372 4.8% 13.4% 31.7% 5.9% 15.8% 37.4%
15 375 60 2414 372 0.5% 9.1% 28.2% .0.6% 9.7% 34.4%
16 375 60 0 433 -00.1% -61.8% -52.1% 1 43.8% -5'V/ 47.0%
17 375 60 0 433 -37.6% 43.3% 55.6% 42.2% 45.0% .52.3%
X111 J. Conclusions from Desorb Adsorb-Desorb Experiment#d
According to the statistical analysis performed by Marty Steinberg of the Corporate
Research Laboratories, color, biosolids and calcium concentrations are all statistically significant
in determining color removal (Koelsch, 1994). It would also appear, based on these results, that
when the soluble calcium concentration was reduced to 30 mg/l, color removal was affected. In
other words, calcium is at or near its minimum concentration at which color removal can still
occur. When calcium or biosolids concentrations are significantly low, weak black liquor actually
experienced large color increases. (See R-4, R-6 and R-12.) Reactor 12 which contained 30 mg/I
calcium and 1207 mg/I RAS had a 39 percent increase in color. This is in contrast to R-1, which
contained 90 mg/I of calcium and 4827 mg/I RAS,that showed 68 percent decrease in color.
Finally, the buffer verses non-buffer results were not significantly different for soluble calcium
levels below 90 mg/I (Koelsch, 1994). This would agree with the buffer experiments run earlier in
the summer which indicated no buffer interference at soluble calcium levels of approximately 54
61

mg/l. These results confirmed the results of the original calcium/buffer experiments which showed
that buffer interference is not a problem at current Canton mill calcium levels.
XIV. Overall Conclusions
Based on the buffer, desorption and desorb-adsorb-desorb experiments, several
conclusions can be made:
1) A significant portion of color removal at the Canton wastewater treatment plant is due
to physical adsorption. Although an exact percentage cannot be calculated, it is apparent that
physical adsorption of color accounts for somewhere between 33 and 100 percent of all color
removal occurring at the wastewater treatment plant. These figures indicate that many of the color
bodies which are adsorbed at the treatment plant are not undergoing any chemical alteration.
Therefore, should the concentration gradient change dramatically in the aeration basins, it is
possible that color could be desorbed from the biosolids back into solution.
2) The phosphate buffer creates significant interference for color measurements at high
levels (approximately 90 mg/l or above) of soluble calcium. Despite the fact that buffer does cause
color test interference at these levels, it is not currently affecting wastewater treatment plant color
measurement. The soluble calcium concentrations at the wastewater treatment plant are well
below 90 mg/l. Should calcium levels at the wastewater treatment plant ever increase dramatically,
some color test interference may occur.
3) Calcium (and possibly other multivalent cations) plays an important role in the color
removal process. As seen in the reactors containing only weak black liquor color, little or no color
62

removal occurs outside the presence of calcium. For this reason, it is important to be aware of the
concentrations and sources of calcium entering the treatment plant.
4) The concentration of biosolids is a significant factor in the color removal process. At
the levels of biosolids tested in the final experiments, the results indicate that the higher the
biosolids concentration, the higher the"percent removal" of color. However, as with the calcium,
at some point additional biosolids would no longer increase color removal.
5) Color removal at the wastewater treatment plant acts primarily on brown color sources
(i.e. weak black liquor). Results indicated that wastewater treatment plant color removal has little
or no effect on color produced during the bleaching operations. Based on this information, color
removal efficiency at the wastewater treatment plant can be judged in terms of brown color
removal (unless significant operational changes occur).
6)Given these experimental results, it is possible to predict the levels of calcium necessary
for different levels of brown color removal. It has been shown that calcium concentration (mg/1),
color concentration (mg/I) and biosolid concentration (mg/1) are all statistically significant factors
in predicting brown color removal. Based on a model (designed by Marty Steinberg) it is estimated
that brown color removal would fall to approximately 0 at a calcium concentration between 35 and
40 mg/l.
In summary, a significant portion of color removal at the treatment plant is due to physical
adsorption of color onto biosolids. This adsorption is dependent on the concentrations of soluble
63

calcium, biosolids and brown color present at the wastewater treatment plant. Should any of these
concentrations be altered, color removal will change accordingly.
The knowledge gained from these experiments has given Champion a better understanding
of the factors which affect color removal. In addition, other mils throughout the country may be
able to utilize these results in designing and conducting localized studies which will aid in
understanding the specifics of color removal at various mills.
XV. Comparison of Results to Past Experiments -
Two studies have been conducted in the past at the Canton wastewater treatment plant that
have a strong relationship with the current study. The first experiments of interest are ones run by
Steve Stratton at Canton in 1986. Stratton's study observed the effects on color removal at the
Canton wastewater treatment plant of switching from calcium hypochlorite bleach to sodium based
bleaching. Stratton concluded that there was "a negative correlation between final effluent color
and dissolved calcium concentration." In other Nvords, as dissolved calcium increased, the final
effluent color decreased. The results also showed that it was possible to increase color removal
when soluble calcium was added. These results agree with the data collected in the current study.
Stratton also concluded that adsorption was the main mechanism for color removal at the
wastewater treatment plant. However, in contrast Arith current test results. Stratton did not find
that the adsorption process acted preferentially on black liquor. The results of Stratton's study
most likely differ from the results of the current study due to the changes made in mill operations
during the Canton Modernization Project.
The second study of interest is one conducted by John Pryately of Champion in 1994.
Pryately used statistical analvsis to correlate color loading into the wastewater treatment plant with
64

color removal across the treatment plant. As mentioned earlier, primary influent color consists of
weak black liquor color, bleach plant color and small amounts of wastewater color from all other
areas of the mill. Based on mill operations. the bleach plant is a relatively constant color source
supplying approximately 32,000 pounds of color per day to the wastewater treatment plant. (This
figure includes both pine and hardwood bleach plant color.) Alternatively, brown color loading to
the wastewater treatment plant can vary significantly depending on variations in pulp mill
operations. Consequently, when color loads to the wastewater treatment plant increase, the
increase is most likely coming from the brown color loads.
Pryately's work divided color loading to the wastewater treatment plant into three classes.
Primary influent color between 0 and 75,000 pounds per day was considered low. Primary influent
color between 75,000 and 100,000 pounds per day was classified as medium. And, primary
influent color above 100,000 pounds per day was considered high. For each of these ranges of
primary influent color. Pryately created a correlation plot with secondary effluent color as the
dependent variable and primary influent color as the independent variable. Pryately found that the
percent color removal occurring at the wastewater treanment plant increased as color loading
increased. Below 75,000 lb/day of color Pryately developed a successful correlation that indicated
that the color change in secondary effluent was approximately -9%. At color loads between
75,000 and 100,000 lb/day, no successful correlation was made. Finally, at primary influent color
loads above 100,000 lb/day there was a 42%decrease in sccondan� effluent color (Prvately. 1994).
The results from this study agree with the conclusions of the current study. As the percentage of
color coming into the wastewater treatment plant attributable to brown color increases, the percent
color removal also increases.
65

XVI. Current Studies
Since the summer work has concluded, Susanne Koelsch and Bill Henderson have been
working with Corale Brierley of VistaTech Partnership on experiments which will measure the
adsorption capacity of the activated sludge. Thus far, one adsorption isotherm experiment has
been run, but more are planned for the near future(Brierley, 1994). In addition, Koelsch is
currently conducting testing to obtain a mill-wide calcium balance. Such information will indicate
which areas of the mill are providing the greatest volume of calcium to the wastewater treatment
plant. It is anticipated that work on the color removal project will progress significantly during the
summer 1995.
XVII. Sueeestions for Further Studv
XVII A. Calcium
The results and conclusions from these experiments lead to a wide array of thoughts on
what new research might be infomtative. One of the first subjects that could be studied further is
the point of minimum calcium for color removal. According to Steinberg and Koelsch's results
this point occurs at or near approximately 35 to 40 mg/I of soluble calcium (Koelsch, 1994). To
test this theory reactors could be designed similar to desorb-adsorb-desorb experiments three and
four with soluble calcium concentrations varying within a range from 25 to 45 mg/I.
In addition to these experiments. reactors could be designed that use other multivalent
cations (for example magnesium) in the place of calcium. It is important to ascertain if these other
cations could be as affective at removing color as calcium if they were found at soluble
concentrations similar to that of calcium. As mentioned earlier, calcium and magnesium compete
with other metals or the hydrogen ion for sorption sites on biosolids. A memo produced by Bill
66

Henderson has suggested that in light of this fact, it may be valuable to substitute copper or some
other metal for calcium and observe the effects. Henderson also suggests using calcium carbonate,
CaCO3, an insoluble forth of calcium, in place of calcium to determine if the form of the calcium
affects adsorption (Henderson, 1994).
XVII B. Bleach Filtrate RecyclerM
One reason to evaluate the abilities of other ions to produce the same color removal effect
as calcium is the mill's recent plans to change its bleaching system. While this color research was
in progress, Canton was announced as the site of the pilot plant for testing a new bleaching process
developed by Champion. The process, known as Bleach Filtrate Recycle (BFRT"'), is expected to
"reduce chlorinated organics, color and other impurities"that are presently sewered to the
wastewater treatment plant. BFRn'will utilize the ODIDOTm technology currently being used at
several of Champion's mills. This process is oxygen delignification combined with 100 percent
chlorine dioxide bleaching. (Chlorine dioxide is used in place of elemental chlorine which is
believed to have more detrimental environmental effects.)
Once the wood chips have been pulped, they are still a dark brown color and must be
bleached. Under OD,00Tm, rather than sending the pulp directly to chlorine bleaching, it is put
through oxygen delignification. This process removes approximately 50 percent of the remaining
lignin in the pulp thereby requiring less chemicals to be used in the bleaching stages. The lignin
removed during oxygen delignification is recovered and burned for energy. In addition to removing
lignin, the oxygen delignification process also removes sonic color from the pulp. leaving it a bright
tan color (Malia, 1994).
Next, the pulp goes through the bleaching process which removes the remaining lignin and
any impurities. In the bleaching sequence, minerals will be removed from wood so as not to scale
67

plant equipment. BFR"m also removes chlorine and potassium from the ash left after burning
lignin. These elements are mixed with water for reuse or treatment. Due to all the recycling and
reuse which occurs during these processes, it is expected, based on lab studies, that bleach plant
organic effluent will be reduced by 50%. Color from the bleach plant is projected to be reduced by
90%.
The question remains how BFRT"" will affect color removal across the wastewater
treatment plant. There are two possibilities. First, because it is bleach plant color that will be
reduced by 90%, BFRT'A may have little or no affect on wastewater treatment plant color removal.
As we learned, the wastewater treatment plant removal processes act primarily on brown color.
Therefore, secondary effluent should have a lower color because less color will be coming from the
bleach plant and wastewater treatment plant color removal will still be acting on the brown color.
However, a second factor must be considered. Because much of the water(and its constituents) in
the bleaching process are being recycled, BFRTm could possibly alter the amount of calcium
entering the wastewater treatment plant. As Koelsch mentioned in her report, weak black liquor
has extremely low levels of calcium when compared to bleach plant effluent and primary effluent.
The two samples of weak black liquor that were sent to West Nvack and tested had 1.9 and 2.5
mg/l of total calcium. The bleach plant effluent samples had 49.5 and 53.5 mg/1 of total calcium,
and the primary effluent samples had 57 mg/1 of total calcium. Therefore, in order to breakdown
weak black liquor color, the brown color must combine with a calcium source. At this point
further research might be useful to determine exactly how much. if any. BFRT"' will reduce the
calcium coming from the bleach plant to the wastewater treatment plant. This information may be
obtained from the calcium balance currently being run at the mill which analyzes calcium
contributions coming to the wastewater treatment plant from the various mill sewers. Presently, it
is believed that much of the total calcium coming to the wastewater treatment plant is from the
68

recovery area of the mill, so a change in the bleaching process may not significantly reduce
primary influent calcium.
XVII C. Biosofids
In addition to completing the tests mentioned above, more studies into how biosolids affect
color removal would be useful. One question which could be studied further is at what
concentration of biosolids does color removal become negligible? Studies could also be conducted
-on-the effects-of-live verses-dead biomass--A study was-conducted by Gary Amy in 1988 on the
effects of live verses dead biomass on the adsorption of total organic halide (TOX). This study
concluded that -'adsorption by dead biomass is equal to or greater than adsorption by live biomass"
(Amy, 1988). A study similar to this one could be run to determine the percent color removal that
occurs when dead biomass is used as the agent of adsorption. It has been noted that adsorption is
affected by cell age. More extracellular polymers are present in older sludge. Therefore, more
metals can be adsorbed (Henderson, 1994). So. studies could examine not only the effects of dead
vs. live biomass. but could also examine the changes in effectiveness as biosolids age.
XVIII. Future of Industry
In looking towards the future of an industry, it is important to first look back and establish
trends. In doing this, it is easy to see that on a worldwide scale the pulp and paper industry has
been moving towards recycling and reducing water use throughout mills. The Canton mill has
been part of this trend. With the completion of the Canton Modernization Project in 1993. mill
effluent was reduced from close to 45 mgd to between 25 and 30 mgd. (This is a reduction of
69

approximately 40 %.) And, now. the mill is embarking on a new project. Bleach Filtrate
RecycleTm, which will attempt to recycle a large portion of the pine bleach plant water.
These trends will not only streamline industrial processes, but they also will allow paper
companies to meet increasingly stringent environmental regulations. The questions arise, "Is a
totally effluent-free mill feasible?"and "How will these changes affect mill waste?". According to
various sources throughout the industry, closure may be possible by the turn of the century.
However, most sources admit that even when "closure" is achieved, spills, leaks and other
unexpected sources of wastewater will occur. A recent article in Pulp and Paper Journal examined
the feasibility of a closed mill. Here, one source was quoted as saying that a greenfield closed-
cycle mill should not cost anymore than a state-of-the-art kraft mill. In must be noted, however.
that he is referring to newly built mills. Currently, the technology is unavailable to successfully
and economically retrofit a mill already in operation. But. he added that a newly built closed-cycle
mill could actually have lower operations costs that a state-of-the-art kraft mill. However, the real
problem may be reliability. In a mill where all the processes are dependent on each other for water
supplies, an upset in one area can have significant consequences on operations throughout the mill
(Patrick, 1994). And, because water is recycled through processes, the concentrations of
contaminants in the water will increase. Such increases most likely will lead to more frequent
corrosion and build-up problems in pipes and equipment.
To deal with the high color concentrations in closed cycle wastewater, new methods may
be necessary for color removal at waste treatment plants. Some possibilities for changes in
treatment may include higher wastewater treatment plant calcium additions, changes in biosolids
concentrations. or even something as advanced as electrostatic precipitation of color (Ochr. 1978).
70

XIX. Conclusion
As we approach the 21st century, the pulp and paper industry appears to have a promising
future. Companies such as Champion are making significant progress in efforts to reduce waste to
protect an environment which is being pressured with ever increasing demands from a growing
population. The Canton color experiments are just one example of how industry and the
environment can both benefit from a better understanding of environmental problems and treatment
processes. Through one twelve-week study it has been possible to identify numerous factors which
affect color adsorption. These results will not only benefit the Canton mill and the Pigeon River.
but, hopefully, they will provide insights for pulp and paper and wastewater treatment plant
operations throughout the country.
XX. Acknowledements
I wish to thank Gabriel Katul, Susanne Koelsch. Derric Brown. Bob Williams and many
other members of the Champion International EOHS and recovery staff. Without their help and
support this project would not have been possible. A special thanks to Susanne for her continued
assistance and guidance.
71

APPENDIX A
Calculation of Color Changes
Percent Change in Color= (Average Color 1987 - Average Color Year X lb/days
Average Color 1987 lb/day
Year Average Color (Ib/day) Percent Change in Color from 1987
1987 346,826 N/A
1993 119,909 — 65 %
I/94-7/94 79,247 — 77 %
72

§ ) \ § / § § § 2 ) § 2 }
�} .
) �:
/ 2k � 2cl
A
\ � )
\ �
)
< \
i mod § 1 ]1V § § §
; gg@9 � R922
) § ) ) dk % ) ) ° ° ` ° A § GRr = �

( ijti }
, K eISS3 , , , 33 (
&
■ ; �i
_
: - _ - - - - - mmE
- - - - - - -
ƒ
/ Ln
§ * } )
Q -
,
/
■ �: $
70
' - - - - - - - 0
_ - - - - ) § \} on
\/ \
�� a
) §
2 ; 2 / m > m ) ! 2 / k ; kl > „ ;;

± e
k �)
) ƒ} GgG2 ; lG7
2 �
; \
® § , ; _ � , � �
j . / -
) 2
rill � \
/ w � � � �
w . a . riee
-0'4—� 522, a§
' w/ 2222 § ! y

['o ro z fA yyJJ ��11 mm ��UU yyJJ mm tm yyyy yyJ 7 zz m ;U �f w N N m ',',dd '',,eD fn zz z 'Uf;
ro to p M D D M D D M D D M D D p M D D M D D M M M D D D D D D M D D D D D M [a
7 In fn + co w + 1J m + m :n ( w J W M J M C/ + + + w fA rn y N [n P !J VJ w CO VJ VJ cn tr
'n tt + ++ `` G+ + z + + U J J In J \ O O O N N ± N N N P P + } } } U N ti y
m z z J G� T In In O O O O O O O J N Nj 'j 'j N U } } } + } U N
n m a rt a n O n n P 9 - o 0 0 3 J J C 0 0 0 "' n
0 o J 'f9+ J •ai -R. 5 y "'1 0 0 0 o M M p M �• 3 M m o '- '" - .-. 000
O O N �.
O D 6 0o J J 0o J ` w J > C7 C7 6 .. .. o
DDDDyd OyG yDD c oin In ooino �' C+
io
n H
rt � �_ E D o -1 rn y o D
51 D' Q T E c C
rt
N N O O �-
:�+ A C A w O O w
V� P ut P D� J 00
'..+ W �^. W - v: U O ID O• N w - w - P P w P w - N O b b -
V• V� V� P A U ,... - - A �- - A - - •^ U J P U - N N N - - - W - -
w
J. A w w J N J 0 - b V� •.+ N w P P w w g a w w w A P J w A N O W :.. O v a O N P P A W W O P b P m
O C
w A AJ •J V� b N - b N - y A 'J� A J lA U O A P J '+/ '+� IJ V� U O b P W P N � � � � W A N - N N P R w
s a
00 W J '-
J "� U x d T Oo C L w W P •"J J �O P Oo d '1 A V� `. A :n IG N C
c w
91U -. - - - - _ v� o 0 0 -- . o � w X
n A V� U A V� C N N W N N A - - •.� U P U U b U U V iU w d ., b J - O IJ O P W - N � �^. P w �D �D b O •"� a w P A N -- - V W N W �: •.I N N - JO c+ a in In :y W U :� '.n J b U O ID W `- N U J C C P A U A W O J N � /�
A O - - O C O O O 0 - -
J - A J J N N J •� b J C U J P A A W N N A U C C N P v 'W P W W W W W A W P In '.+ N •W P w w ,
A N U U W N A
V. U O O w -- N w P w U C O C O O U U A N O w - V. 'p. - A W O J d •w 0� P m
C O - "' T O J O - .•� C '.� P - .... _. N ... O N - - - - .... - ... - �. - C - '..I
•W J .- w N � IJ - N - N �-• J '- "'• J W O J � 'V
b w d N C` C - P O a to U O. U 7- O O O O O
d J J J P P J A W P W W P A P P v� w P P w A G c " W D P g W tr W J A N w U W •.� •w •y - J A A A A A .- I
J
w - W W w U J G P V• V� W O O O w J A J
^ n -• C O '- O O C C C C C O O O O O O O O C C - C - O - 0 0 '- O C O G O - cc 0 O O O P
v� V� O O � y Wy A J W T LI N - ? N C - b P w P U b P J U - •Ay O yy A Ty Ny J N y N} O
V. Z P W U d Z N d O� } T •'.1 w ".l j O ".l O.

m
p
ry v P V wi T N r N 1� N C P a0 N Q - N p� Ip OO a r,
O O 'n — O — O vi �n a. �D ^. 1� N m
- N
T C0 'D
W OHO
O •-• •-• ^: - O O�
O T �O �D
O e O O O C C O _ rO IR OO_ O O O O G
'V I� O O - � OR
r
C W M r. -n F
OO 06 GC ,6 vt Q P Q Q T
J CT V' t� P V' Cd W r, b h r a P N C - r T v 'JO _ = Gl T •-. N T ^ r. Q P - t!.
O C 0 0 O C C C tV e N O N
p -�
a
�p O e e C !2 66 0 0 --. O O C O C
y o o a o o F y F - - v 0 0 9 0 9 u
p0, oF' yFv� Tl"' v� � „' — NN v vv 6_ FvFvoiQ 'c
f2 0 . v � " ao
N FP ¢ c� �N �rvaaAA o 03 � '00. aQa6Q Q ¢ a0, 0
P ..
F `. 0 0 a ... c o 0 0 o N
m v Q ¢ < o u aQQQ Qp00.0 oo co F"n oQ P.
0 oa- QQ y -' y 5 oom P. rm
+ 0 00. 0. 0. 4U y m 0 o p0, 0.0 co w v h t� n w � U "z � � � U II
.-1 13. 0 00. 0. UU � � ;n •r, � n + � � r � + U
paN¢ N0. 0. UU0. UU1ppa�7 + + ry ddo rv¢ h + + UUp�aaa (� ppa�1pW� 0
3r1 aW wroaP133wwa � aQi33aw vw 2a xaawam 3333

REFERENCES
Amy, Gary L., Curtis W. Bryant, Bruce C. Alleman, and William A. Barkley. 1988. Biosorption
of Organic Halide in a Kraft Mill Generated Lagoon. Journal WPCF. 60: 1445-1453.
Bach, Orville and William Barnett. May 9, 1987. An Economic Impact Analysis on the
Recreational Benefits of a Restored Pigeon River and A Financial Analysis of Champion
Intemational Corporation's Ability to Provide for a Clean Pigeon River. Walters State
Community College.
Bach, Orville. October 25, 1994. Personal Interview.
Brierley, Corale L. November 28, 1994. Biomass/Metal Techniques. Memo. VistaTech
Partnership, Salt Lake City, Utah.
Carolina Power and Light. May, 1994. Hydropower Facility Site Tour,North Carolina.
Carpenter,William L. March 14, 1986. Color Measurement Procedure. Notes to File. National
Council of the Paper Industry for Air and Stream Improvement, Inc, Gainesville, Florida.
Champion International Corporation. April 15, 1994. Discharge Limitations: Canton Mill.
Canton,North Carolina.
Champion International Corporation. August, 1994. Datastream Calculations of Discharge
Averages. Canton,North Carolina.
Champion International Corporation. August, 1994. Datastream. Database.
Champion Intemational Corporation. 1994. Fact Book 1994. Stamford, Connecticut.
Champion International Corporation. NPDES Permit Color Compliance Requirement. State Line
Color Calculation. Canton,North Carolina.
EA Engineering, Science and Technology, Inc. February 19, 1988. Synoptic Survey of Physical
and Biological Condition of the Pigeon River in the Vicinity of Champion International's
Canton Mill. Sparks, Maryland.
Eckenfelder, W. Wesley. 1966. Industrial Water Pollution Control. McGraw-Hill, New York,
New York. Pp. Pages
Clark, Alan. 1993. French Board River- Basinwide Plan. Draft. Environmental Management
Commision, Raleigh,North Carolina.
Henderson, Bill. December 14, 1994. Notes from the Metal-Microbe Interaction
Presentation/Teleconference. Memo. West Nyack,New York.
Koelsch, Susanne. 1994. Effluent Discharge Graphs. Champion International Corporation.
Koelsch, Susanne. October, 1994. Personal Conversation.
78

Koelsch, Susanne. September 23, 1994. WTP Color Removal Mechanism Studies Status Report.
Memo. Champion International Corporation, Canton,North Carolina.
Lang, Edward. Properties of Color in Kraft Mill Effluents That Affect the Making of Mill Color
Balances. St. Regis Paper Company, Cantonment, Florida.
Malia, Peter J. 1994. Closing the Loop at Champion Pulp and Paper Mills. Special Report.
Champion International Corporation, Stamford, Connecticut.
National Council of the Paper Industry for Air and Stream Improvement, Inc. December 30, 1971.
An Investigation of Improved Procedures for Measurement of Mill Effluent and Receiving
Water Color. Technical Bulletin Number 253. New York, New York.
National Council of the Paper Industry for Air and Stream Improvement, Inc. February, 1984.
Effect of Biologically Treated Bleached Kraft Mill Effluent on the Periphyton Community
in Southern Experimental Streams for 1976 to 1977. Technical Bulletin Number 421,
New York,New York.
National Council of the Paper Industry for Air and Stream Improvement, Inc. March, 1984.
Effect of Biologically Treated Bleached Kraft Mill Effluent on the Periphyton Community
in Southern Experimental Streams for 1979 to 1981. Technical Bulletin Number 423.
New York, New York.
National Council of the Paper Industry for Air and Stream Improvement, Inc. July 31, 1970. The
Mechanisms of Color Removal in the Treatment of Pulping and Bleaching Effluents with
Lime. 1. Treatment of Caustic Extraction Stage Bleaching Effluent. Technical Bulletin
Number 239. New York, New York.
National Council of the Paper Industry for Air and Stream Improvement, Inc. December 28, 1970.
The Mechanisms of Color Removal in the Treatment of Pulping and Bleaching Effluents
with Lime. 11. Treatment of Chlorination Stage Bleaching Effluents. Technical Bulletin
Number 242. New York, New York.
National Council of the Paper Industry for Air and Stream Improvement, Inc. June, 1994. Human
Perception and Biological Impacts of Kraft Mill Effluent Color. Special Report Number
94-07. New York, New York.
National Economic Research Associates, Inc. February, 1988. Benefits and Costs from the
Reduction of Color Effluent from the Champion Mill into the Pigeon River.
Nichols, Robert, David Moreau, Stiftel and Hyman. 1988. "A Review and Analysis of Fourteen
Environmental Assessment Methods." Combinine Facts and Values in Environmental
Impact Assessment. Pp.] 15-223.
Oehr, Klaus. February, 1978. Electrochemical Decolorization of Kraft Mill Effluents. Journal
WPCF. Pp. 286-289.
79

Patrick, Ken L, Jim Young, Kelly H. Ferguson, and Andy Harrison (editors). March, 1994.
Closing the Loop: the Effluent-Free Pulp and Paper Mill. Pulp and Paper. 68: S I-S24.
Tsezos, M and W. Seto. 1986. The Adsorption of Chloroethanes by Microbial Biomass. Water
Research. 20: 851-858.
Volesky, B. 1990. Removal and Recovery of Heavy Metals by Biosorption. BiosoMtion of
Heavy Metals. CRC Press, Boca Raton, Florida. Pp. 743.
United States Environmental Protection Agency. November, 1993. Proposed Effluent Limitations
Guideline and National Emission Standards for Hazardous Air Pollutants for the
Production of Pulp, Paper, and Paperboard -Industry Subcategory Definitions. EPA Fact
Sheet. EPA-821-F-93-004. North Carolina.
United States Environmental Protection Agency Region IV. September 25, 1989. Authorization to
Discharge Under the National Pollutant Discharge Elimination System. Permit Number
NC0000272.
Water Pollution Control Office of Tennessee Department of Environment and Conservation. April
5, 1994. Champion Paper- Update. Memo.
Weber Jr, Walter J. 1972. Physicochemical Processes for Water Quality Control. Wiley-
Interscience, New York,New York. Pp. Pages
Wilson. Dawn. June 15, 1994. Cleaner Pigeon River Expected. The Mountaineer. Waynesville,
North Carolina. 71.
Yost, Thomas B. February 12, 1992. United States Environmental Protection Before the
Administrator in the Matter of Champion International Corporation, Permittee. NPDES
Docket NumberNC0000272. Pp. 1-18.
Ziobro, George. 1990. Origin and Nature of Kraft Colour: 1 Role of Aromatics. Journal of
Wood Chemistry and Technology, 10: 139-144.
Ziobro, George. 1990. Origin and Nature of Kraft Colour: 2 The Role of Bleaching in the
Formation of the Extraction Stage Effluent Colour. Journal of Wood Chemistry and_
Technology. 10: 151-168.
80

Pigeon River
A Survey of Pigeon River
Re-introduction Efforts
Joyce Coombs,Virginia Harrison,J,Larry Wilson ✓
University or Tennessee
Department of Forestry,Wildlife and Fisheries •.�` � �4:,:.,
Jonathon Burra --
Tennessee Department of Environment and -
Conservation,Water Pollution Control
Knoxville,Tennessee
• Pigeon River has suffered historical
Management Goals degradation from Industrial discharge.
• UT goal:to collect,tag,and track • Water quality Improvements have led to
the return of some fish species.
re-Introduced fish species
• Joint re-introduction project by TDEC-WPC,
TWRA,TVA,USGS,USFWS,NCWRC,
BRPP,CFI and UT -. .
• Stale goal:re-Introduce native species as • '�
part of efforts to restore river ecosystem
health and designated usages
Pigeon River Pigeon River
1983 1995
rZ`tiz,,
E;>
N
1

Pigeon River, Newport Annual Mean Effluent Color(lbslday)
February 2002 Including Permit Limitations(Ibs/day)
9882003
am
B�®
y d i✓x*aekxk'„�`3 r
g;w®
m f•'M:>wiarut r.Ya
Native Fish Collected 1988-2001 Pigeon River:
Modemizalion What Species are Missing?
bluebmaet darter gm dartor
tangerine dader can,darter
N pDenton river it it r wounded darter
'a aldpetZIrler longhead darter
■Tannery
y blotchslde logperch blueslde darter
6
N wamalnt shiner siriyud shiner
G mimorehi minnow
mimic
slargead minnow mlmlcehlner
! bullhead minnow mountain
chub
E cpemnchub mountain medium
l noMem etudeah mountain brook lamprey
Z '�, American brook lamprey mooneys
Yea s
Pigeon River:
Mollusk Populations Tagging
• Approximately 40 species of news mussels are belleved to
have thrived In the Pigeon River historically.There are no ,ate
native unfonid mussel species populaduns.
• Several species of river anails were m4ntroduaed in 1996
(Leptoxls six.,Pleumcera sp,Elfmfa sp.)and 1999 go sp., " L�
Campefoma sp.,UMasia sp.);Leptoxis sp.and Pfeurocom up. p.L
are reproducing well
t<.
Ra4ntr ,iducgon of mussels began In 2gg0.Survival of sorne
mussels has been observed but mproductlon is lndetenNnato
at this time.
2

Tagged to sp. Spiny Riversnail
to f/uvialis
Pigeon River: Fish Consumption Advisory
Trend Summary and Current Status
• IBI scores have risen from' oor'In late e0'9 to `WA �NG CARP
P -CATFISH r ",.
'good'range In 2000
• Strong Improvements In species diversity.
abundance•and sport fishery :fmm.thla,nosy of w6ar co"Wn miwM nw st:
klmb thouqht to^.incnus pr'rp4.v1 tenor
• Fish Consumption Advisory removed entirely. othor uriwe IlYwo7ln�humena
• Agencies began redntroductions of native fish These fish sliauld not 6'e4n.Alto childi''eh,
regnant or nursing wo an NI l ere shaald
species In March 2007 itsonsumpbenahtna emealper man
• A total of eight species re-Introductions orpuarnvrt>F' Mo
Re-establishing Native Fish Initial Candidate List
2001-2002
Criteria for selecting candidate fish species
to be relocated to the Pigeon included:
• Gilt Darter—Percina evides
• Historic range • Bluebreast Darter—Etheostoma camurum
• Available habitat • Blueside Darter—Efheostoma stigmaeum
• Ease of collection and transport • Stargazing Minnow—Phenacobius uranops
• Available In large numbers nearby - Mountain Madtom—Nofurus eleutherus
• No listed species
3

Gilt Darter Bluebreast Darter
Percina evides Etheostoma camurum
AY
Blueside Darter Stargazing Minnow
Etheostoma stigmaeum Phenacobius uranops
Mountain Madtom Methods
Noturus eleutherus
Collection by seining from Little Pigeon,
Nollchucky,and French Broad Rivers
• Tagging with various colors of Visible
;3 Implant Fluorescent Elastomer(VIE)
",1�g Snorkeling surveys to locate the tagged
fish and YOY
1-!=r em-"W
Ink

Collecting Fish: Field Tagging
Nolichucky River
VIE Tagging
Injecting VIE . _
_r
Transporting
VIE Tag Under Blue Light
f �w.

Acclimation and Release
2002 Snorkel Survey
2002 Snorkel Survey 2002 Snorkel Survey Results
• Observed 175 fish,a density of 58.3 fish
In 900 m2 or 19.4 fishN00 m2
• Primarily Gilt darters with few bluebreast
darters
• Representatives observed from all 4 batch
re-Introductions of gilt darters
Phase 1 Observations 2003 Project Activities
• Re-introduction of selected darters to the Relocated 3 new species for 2003:•
Pigeon River appears viable where adequate habitat Is Identified American brook lamprey,mountain brook
lamprey,and stripetail darter
• In the lab study,VIE tag retention was
good for 100+days and did not cause • Relocated additional common snails
mortalities
• Monitored survival,movement,and
• In the field study, VIE tags have been reproduction In tagged individuals
retained over 2 years
6

Collection Results Collection Results
Tagged and relocated as of 10/28/2003: Relocated as of 10/28/2003:
Blueside darters 766 Stargazing minnows 270
Gilt darters 1015 Mountain madtoms 381
Bluebreast darters 334 American brook 477
lamprey
Stripetail darters 493
Mountain brook 239
lamprey
American Brook Lamprey Mountain Brook Lamprey
Lampetra appendix khthyomzon greeleyi
�t
Stripetail Darter River Snails
Etheostoma kennicotti
P/eurocera sp. Leptoxis sp.
7

Pigeon River: Snorkel Survey Results 2003
What Species are Still Missing?
bluebreast dartedarter
ollgil darter
Observed 102 Gilt darters during random
bluab ne Carter ellt ddarter g
dverdarter wounded darter transacts of 21 habitat sites
MlImeil darter lonphead dader
blotchslde loepemh blueelde darter —18 tagged,84 untagged
wamalnt shiner aW,,td ahlner Gilts were located at seven sites including the
mldorshlner saffron ahlner redntm site
alamatlne minnow mlmle ahlner Three genera of river snails(to,Lepfoxis,
bullhead minndw blatched Nub epolen chub mountain medtom Pieurocera)were located at five sites:
-3%TI•••,+1—,+11•,+18•Denton-*)
atudnan mountain bmol�umprey Successful recruitment for Le foxis and
nmedran brook lamprey mooneye P
2ee1,2M,mm Pleurocera
Movement and Recruitment
` Pigeon River Re-introductions
.} F
P N".r�:•+tArlr ;r . „1.�: Year Organism Genera/Sp. Individuals
F!�ryw,.s�t -fix 1996.2003 Snail 8 80k-80k
2000.2003 Mussel 9 145
i1 S ;�, 2001.2003 Darter 4sp. 2808
2O02.2003 Madtom 1 sp. 381
2002.2003 Minnow 1 sp. 270
rJte thl - 2003 Lamprey 2 sp. 716
2004 Project Plans-TN Tangerine Darter
Percina aurandaca
• Continue relocation of current species
• Relocate additional common snails
• Continue to monitor survival,movement,
and reproduction in tagged Individuals
• Begin propagation of the tangerine darter
• Analyze aquatic Insect data Y-r~`� '•sR►,r• -`'"'
8

2004 Project Plans-NC NC Candidate List
• Complete habitat assessment of NC reach of
Pigeon River • Silver shiner-Notropis photogenls
• Identify candidate species and sites for • Saffron shiner-Notropis ruhricroceus
re-Introduction
• Identify sources for re-Introduced species • Telescope shiner-Notropis tolescopus
• Analyze aquatic Insect data • Tangerine darter-Porcine aurantiaca
• Golden redhorse-Moxostoma erythrurum
Questions?
9

Presentation to the Technology
Review Workgroup
Tuesday, December 16, 2003
Blue Ridge Paper Products Inc.
Canton, NC

Agenda
• Welcome/Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Bob Williams
• Pigeon River Re-introduction Effort. . . . . . . . . . . . . . .Dr. Larry Wilson
• Environmental Capital Plan/Financials. . . . . . . . . . . . ... . . . ..Bob Shanahan
• 2001 Permit: Color Reduction Improvements... . . ...Michael Ferguson
• Color Performance. . . . . . . . . . . . . . . . . .Melanie Gardner and John Pryately
• October 2003 Report. . .-. .. . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . ..Bob Williams
• December 2003 Report. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . ..Bob Williams
• Questions
• Lunch
• Tour

Monthly Average Secondary Effluent Color (lbs/day)
Jan-88 through Nov-03
700,000 — ---— --- ----- --- ....-.. -- ---- - —
600,000 --- ----- ---- ----
500,000 - -- — -- -- ------
400,000 -----
300,000 - -
200,000 --
100,000 111110,
0
o? m ao ao rn rn M T rn T T T T rn T rn rn rn rn T T rn T rn o 0 0 0 0 0 0 0
C j C C C C C j C C -) C j C j C C j C = C C j C
(6 (0 (4 -, 16 (0 ' f6 -, f6 (6 ' CO ' (6 --) l4 f6 f0 cD (6

Pigeon River Color Upstream of the Mill and
at the State Line
Monthly Averages from Jan-88 through Nov-03
250.0 —---- -- —- --— -- --
200.0 -- -
150.0 --- -- --
c
2
`o
0
100.0 —
50.0
0.0
M LO (O (O r- r- N CO m m O O N N CO (7
m m m m rn m m rn m T rn m m m rn rn m rn rn T m rn rn o 0 0 0 0 0 0 0
--#--Upstream Color ITN/NC Color

Pigeon River Color at Hepco
Monthly Averages from Jan-88 through Nov-03 -
350
300 -
250 — -----
`0 200 —
6
U
m
F-
150 ---- -
100
50
0
00 \C§) 00 \00 00 \cp O^ �9^ C�' �0g' 00 �00 Off` �(�K 00 �00 00 �00 OI \c1 00 �qb O°' �O°' 00 �00 O1 �O^
)J )ate )J )ate )J )ate )J )ate )J )ate )J )ate )J )ac )J
t Hepco True Color

Environmental Capital
Plan/Financials

Blue Ridge Paper Products — Canton Mill Regulatory Impact
Regional Hazeluc .uuna.
VISTAS -TRI S✓ 5(/ f°
-Lower Threshold
Global Climate y6atrrCGt �h.lo✓t�e{D ��
Change Title V` PM/Ozone
NAAQS. .
Nox Controls
Industrial zM
Combustion MACTxm
Crj
P
ANTON '
V
L4BMp
��
RCRA
Cluster Rule & r
MACT
Landfill Rules.
TMDL )7
Cooling
WQS A Wate�r4ntalce .
Compliance Maintenance - $14 Million

2001 NPDES Permit
Color Reduction Improvements

Table of Color Deduction
Measures
M-6rRCdueho ENE a um 22001 2002 2003- 2004 2005 2006
1 2-Hour Color Testing
2 Iiebergott Reconuendations I Iermnted - —
3 Use of Off-line Clarifier -
4 Hwd Brown Recovery Tank Line to Pine Blow Tower
5 I-Hour color testing before and during nmintenance outages
� t _
6 Installation ofMechanical Seals �-
7 Improvement of equiprrrnt used for handling Pine and
Hardwood Knot Rejects
8 ISI.5M spent on Bleach Filtrate Recycle Improverrents _
--—
9 Installed P¢teBrownstockControl Logic irrproverrents
10 Detemvned nultiple contributers and interaction effects u
causing Sewer Generated Color
r
11 Bench-scale sewer-generated color work —'-
12 Process Optunization/SixSigrm Teamdesignated for 1 -- -- --
Hardwood Fiberline color improvenents
13 Improven-ents rnade to Pine Brownstock sumps for better
recovery W .
14
White Rot Fungus Trial-growth efforts unsuccessful,rmy
re-visit in future.
15 Pure Brown Recovery Tank Line to Hwd Blow Tower
16 De-WateringlDisposal of Pine and Hwd Screen Rejects „
17 Pilot Scale Plant for Ozone De-Colorization --
18 Implerrent full-scale Ozone Decolorization systemifPilot-
scale study is successful

a
lf� L "�-`�� -t .yyµ� Ta� +'y' � �{,.�{.,3,�{�}F V�My�)3y$yT .M4fYK.� �a"ry fr +6 }4�{•'yl�.°I4 7�4�. kifHf3 K�'°' ��,g611 � � rl
� 1. � b SN ! [f/ r �t } } 1 N�
� !' 0 t :i a Rf 0Sfl4A Yx F J aK
I Got
� H4fi S+33A z+ t
+f +a. n,� 'F ' "Ey r.k ,`e n �,nV1r, i , y� . 1 v..� !i -;`+ ! r Nf +S f
•v,°`vLA ati. " '"• r�µs"�����yy ((}}� .��{jLiv Py� x� ��_'���yy
PC0494PSIG
d . ;
2.5
fi`��5 v .aK�� '• %eA ? w" 9.v" `. c OT' s ail 0 1 �'v r„ ',4#.. °+ •°Ss+''H wi
7++f n r " 'ten 51 tx+ ,ma 3�sY�axre r4 = AUTO Ole
4
ER 83�$
J lj%, p*�, v a ;y},w
V K,�VIY�}.^1TS7raR �YY ` e"°° °%• ' «a' s.� a
i B68M STdG
"ATOd�P 12ND STG SllkP T
�e r216 am"O a s 1 SStNENK 1�4ii1H0
0386
SPILL TAFBt LEVE
�` � ,`a•° LC0495
m ;FCT 17.0 7
? FC0496 '
` r GPM —1 . 0
LOC 100.
r ` BROWN'STK =OTO 100.0
_w SUMP PUMP
T'M4
v.y •:;a '•, s°"gyp Je
e „
,

BFRTM Improvements
• Greater than $ 1 . 5 MM has been spent to improve
BFR over the last 2 years
• Three phases of improvements:
— Piping
— Media Filters
— Ion Exchange Softeners
• Improvements completed in October 2003
— Just now beginning to see true benefits
• Goal: To get rid of the individual component
liability and reach 100%o Uptime with the
exception of planned outages.

a
44
NI
IhLaul
1 ' 1.�,. ., _.� ,� ,I, is —' ; i.� �I,,•� i _ +�'`'� �±
!
.x4k'

BFRTm Improvements Continued
J W y
Filter — allows for
o VNew North Media
three Media
Filters to
< •� 1 .i l remain
requires repairs.
-a Elul � i. A i•'
if
jz ��6A a f E

BFRTM Improvements Continued
Top view of three
original Media Filters, = rc=
which have all been
replaced to reduce f:
v E^
MRP downtime and
d
improve throughput. - �r
f
Ilk

MRP Uptime
A:2
100 ` -j
90
80
70
60 _
50 Uptime01
4017,
-
30
20
10
0
00
��, Q�� C P ��, , off, G'` oJ,
�` 5

s
Alh
-
IL
G
�` i p' Yt 1 �.� �•�%+' ATE. � � ,
6,
yy1

pp
i r t.A
^�
p
,Ja" %'" ��3s " _'� `•, �I• max ' °,
t �y{ FFyy �n
+' a �� 3 ,nvgrou
,1 3 C x
��� � — ` u'i'♦ �....� �L�.� * 'IBM .+F"' It b4 il'4 j I.�
"rQ f
A'

Improved System for Handling
Pine and Hardwood
Knot Bin Rejects
-
Hardwood Knot Bin
grating prevents
y knots from entering
rain but allows
rti� f r� filtrate to flow to
4
L �
dedicated sump
4 :fir nomr joy Y1� i� tik „M,
�—i-t.3. k'x'.21#f��w �' A� f �
F� a; pump for recovery to
r } the Hardwood Spill
-Tank.

Pine Knot Bin Sump Pump
lz
a .x3q ffr'k
�
� r
I
e P

Offline Spare Clarifier
�11 Iq �'� X v 8 er'`">✓"'�, f sf .� .{ 4 a i �
a i�.� 3 ' ��y�?tA � _." { r, �� sty`° �•b � s'�- ^ a i" �� �y � 1^ w.
� ^..:,."rM t s.v �i, � iiiu".°!31 £ �-5 _ r 1' T Y4+�.'r _ t{•s<.��'§�su�„a ."i �i,r.�as� �s �1-�n
r21T�F.a_`
.., r
f
kw
r
•' A- 1° i I M1 LY �� GL Y 1. A s'i
'X}�� } Ly L r
..#i"`*T IY£uia: w.� 1 ii°''rm Rk ]"SA ? arsp: •�v.�s �, x.;..
,4� .-, ".' ., xS:�* M ,«>n� •ras ..., '�s4. yu,t, 7yas S+e, a..f.�.
R�

Installation of Mechanical Seals
• Customized Double
Mechanical Seals and 1 kF{ it
Water Management
Systems have been ;..
installed on the 18
digester re=circulation Y_
pumps for clean water
reductions in the
Digester Courtyard.

Increased Color Testing and
Improved Communications
iM
Around-the-clock 2-hr
color testing on the
Primary Influent
�� k. r " C� y. .Ayl� � t°a`4• � .gyp Fvy sM J7
Color testing every
" ' ' i`� � hour before, during
an upon start-up of
+ f `
outages
WWTP Crew leader
ye . communicates
elevated color results
to Pulp Mill and
Recovery Foremen
�y Hk

Color (lbs/day)
N W pA (n m
o O O O o S
COO O O O O O
Jan-01
I
Feb-01
I
Mar-01 I
Apr-01
I
May-01
Jun-01
I I I
Jul-01
I
Aug-01
Sep-01
Oct-01 00
Nov-01 =r
Dec-01 0 `<
Jan-02
Feb-02
Mar-02 o CD
Apr-02 3 CD
C
M ay-02 O
Jun-02
Jul-02 Z
O
Aug-02 c m
Sep-02CD
w
Oct-02
N ov-02 0
O
Dec-02
Jan-03 O
Feb-03
M ar-03 I
I
Apr-03
M ay-03
Jun-03
Jul-03
Aug-03
Sep-03
0 ct-03
Nov-03

Average Sewer Area Color by Year
1997 - 2003 (through November)
35000
30000 12Note:for2002, 5,000lbs/day was included for
contribution from Quaternary Screen rejects.
For 2003, 3,000 Ibs/day contribution was
25000 included.
a
0 20000
.n
0 15000 _
o
10000
5000
ra
fi i4" Fi' Lbts; j..
1&2 FL's D1 + 513- Recovery, 3A- No. 1/2 Eo, 2B - Digesters, Unaccounted PWs- 11 & 12 Contaminated Combined
Pine D2: BLO,CRP No.2FL BSW, No. 1 FL Condensate Condensate
02 Delig
0 1997 p 1998 131999 02000 02001 p 2002 p 2003

5.
1997 Average Measured Color in Mill Sewer Areas as a Percentage of
Primary Influent Color
Annual Average PI Color = 87,485 Ibs/day
Annual Average SE Color = 62,318 Ibs/day
0% 12% g] 2B -Digesters, No. 1 FL: 10,342 ppd
i
w s 3A- No. 1/2 Eo, No.2FL BSW, 02 Delig:
fi
14,122 ppd
390 16%
" ❑ PM's- 11 & 12: 5,120 ppd
7
c
T
w
Uq 513-Recovery, BLO,CRP: 10,871 ppd
6%
1&2 FL's D 1 + Pine D2: 13,407 ppd
12% ❑ Unaccounted: 33,601 ppd
15%

2003 Average Measured Color (Ibs/day) in Mill Sewers as a Percentage of
Primary Influent Color
PI Color = 56,568 Ibs/day
SE Color = 45,192 Ibs/day
7% 1%
15%
3%
r
3 ❑ 1&2 FL's D1 + Pine D2: 8,196 Ibs/day
H 5B- Recovery, BLO,CRP: 12,074 Ibs/day
, p 3A - No. 1/2 Eo, No.2FL BSW, 02 Delig:
10,196 Ibs/day
❑27% 2B - Digesters, No. 1 FL: 2,593 Ibs/day
23% g Unaccounted: 14,270 Ibs/day
NF
- ❑ PM's- 11 & 12: 1,824lbs/day
E Contaminated Condensate: 3,956 Ibs/da
{
❑ Combined Condensate: 459 Ibs/day
19%

Historical Sewer Generated Color
Studies
• 1995 Study was performed to determine if
sewer generated color is removed across the
WWTP
— Average SGC for Pine Dlwas 32-47%
— Average SGC for Hardwood D1 was 64- 88%
— Pine filtrates increased in color an average of
21 % across the WWTP
— Brown source color removal ranged up to 70%
across the WWTP in this study

Historical SGC ,Studies Cont.
• 1994 Study was performed to determine
what type of color the WWTP is removing
and what factors are significant in that
process .
— Brown color sources were removed across the
reactors by an average of 62%
— Bleach plant color increased across the reactors
by an average 0
f 3%

Decent Benchscale Sewer Generated Color
Work
Pine D1 Filtrate
Simulation of In-mill Sewer Mixing and
Resulting Sewer Generated Color
Samples taken 11/30/03 - 1218/03
3500
pH was increased to 11,then brought back to 7.6.
3000 Color measurements were made at each incremental pH.
2500 —
0 2000
t` 1500
Original Sample
1000
500
0
2.75 5 7.15 9 11 9 7 7.6
pH

Benchscale Study ' 1
Pine D1 at ph 9
111aa
11
111
CL
CL
I
11
1000
v i. evl� vevg � v . �vv} 6e� • 'Ev aY' A*�u' b
11
SY
1 '.>.;1'.,a.s_:�` i..': .. .. tip,,=`s. •.. ,. ° ... °5-... .�'»,_
00 00 11 00 0 11 11 11 0 00
Date Sampled
pH Decreasing

Pine D 1 sample at different pH ' s
AAa 3 Afi+ .f 3 }`S�t 4 N Sa X
wpw �t _
fli
�7` t'aY i iw`ar t 4 � Mfill
is
aZµ3Y
M'
I
Original Pine D 1 filtrate, pH = 2.7 Pine D 1 filtrate at pH = 10.9, filtered 4 times

October 2003 Report
• BRPP recommends an annual average
Secondary Effluent color limitation of
42,OOO lbs/day
• BRPP recommends a monthly average
Secondary Effluent color limitation of
52,OOO lbs/day

December 2003 Report
• Current strategy to remove additional 3 ,000
— 8 ,000 lbs/day is. to :
— De-water/Dispose of Quaternary Screen rejects
— Pipe Pine Brown Spill to Hardwood Blow
Tower for additional capacity
— Process Optimization
• Hardwood washing
• Sewer Generated Color

Proposed ozone Color Reduction
Treatment
• Lab study showed 70 % color reduction of Hardwood
Eo/CRP Purge mixture is possible. Pilot scale study must
be conducted to determine overall feasibility and true
effects on mill effluent color.
q
I.
t� 4 M1 y r
Y t'
r.
i9, y
d

Questions ?
Lunch
Tour