Submitted to U.S. EPA
August 14, 1996

Table of Contents

Executive Summary

Section 1:
A Review and Assessment of the Ecological Risks Associated with
the Use of Chlorine Dioxide (available as a PDF document)

Section 2:
An Assessment of the Formation of 2378-TCDD and 2378-TCDF
when Chlorine Dioxide is the Oxidizing Agent in the First Stage of
Chemical Pulp Bleaching (available as a PDF document)

Section 3:
Eco-System Recovery: Liftings of Fish Consumption Advisories for Dioxin
Downstream of U.S. Pulp Mills

Section 4:
Comments re: Environmental Benefits of Extended Delignification

Section 5:
Comment re: Relationship Between AOX Discharges and Formation of Dioxin and Furan and Polychlorinated Phenolic Compounds

Section 6:
Comments re: Encouragement of Implementation of Totally Chlorine-free Technologies

Section 7:
References

Executive Summary

Improved Water Quality

Section 1 of this submission, "A Review and Assessment of the Ecological Risks Associated with the Use of Chlorine Dioxide," was prepared by a panel of esteemed scientists and submitted to the U.S. EPA at the February 10, 1994 public hearing. The report clearly demonstrates that complete substitution of chlorine dioxide in the first stage of pulp bleaching provides considerable environmental benefits.

The panel of scientists reached a unanimous opinion. Based on the existing available data, the panel concluded that chlorinated organics from mills bleaching with chlorine dioxide, employing secondary treatment and with receiving water dilutions typical of most mills in North America, present an insignificant risk to the environment (Solomon et al., 1994). This opinion was reached independent of the extent of delignification of the unbleached pulp prior to bleaching.

Elimination of Ecologically Significant Bioaccumulation

The panel of scientists reached a unanimous opinion with the following conclusions:

• The examination of analytical results for laboratory and mill bleaching, the latter comprising 163 samples from 9 Canadian, 6 U.S., and two Swedish mills where ClO₂ had completely replaced Cl₂ in the first stage of bleaching, showed that 2378-TCDD was not found in either mill effluents at detection limits ranging from 0.3-0.9 pg/L (ppq), or in bleached pulp at detection limits of 0.1-0.3 pg/g (ppt). Therefore, complete replacement of Cl₂ with ClO₂ in the first stage of bleaching results in virtual elimination of 2378-TCDD.

• At 6 of the 17 mills mentioned above, 2378-TCDF was not detected in either mill effluents or in bleached pulp. At 5 of these mills, 2378-TCDF was seldom detected, while at 6 of the mills, 2378-TCDF was detected frequently. At the mills where 2378-TCDF was detected, the most probable cause was elevated amounts of dibenzofuran (DBF) precursor in the unbleached pulp. However, in a few cases, sources external to the bleaching process could not be ruled out.

• Downstream of mills which have completely replaced Cl₂ with ClO₂, concentrations of 2378-TCDD and 2378-TCDF in fish were found to be declining rapidly. More importantly, in the only case known to the Panel of a new bleached kraft mill which has never used Cl₂ and only used ClO₂ for bleaching, after one year of operation, 2378-TCDD and 2378-TCDF were not detected (at 0.1 ppt) in fish caught directly downstream of the mill.

• Taken together, the available data, although limited with regard to specific ecological responses, support the conclusion that at bleached chemical pulp mills, where ClO₂ has completely replaced Cl₂, no adverse effects caused by 2378-TCDD or 2378-TCDF from bleaching should be expected from mill effluents.

N.B. This opinion was reached independent of the extent of delignification of the unbleached pulp prior to bleaching.

Elimination of Fish Consumption Advisories

The report, an analysis of the U.S. Environmental Protection Agency's National Listing of Fish and Wildlife Consumption Advisories (NLFWA), reveals three important findings regarding fish consumption advisories for dioxin. First, the number of waterbodies under a dioxin advisory represent 2 percent of the total number of affected waterbodies in the U.S. According to the EPA, there are currently 1,740 waterbodies under some type of advisory restricting fish or shellfish consumption. Of this total, 37 waterbodies in select pulp and paper states are under a dioxin advisory.

Second, of the 37 waterbodies with a dioxin advisory, only 18 are downstream of U.S. pulp mills, amounting to less than 2 percent of the total 1,740 affected waterbodies.

Third, the small number of waterbodies under a dioxin advisory is steadily diminishing. Since 1991, 13 states have lifted a total of 17 dioxin advisories from waterbodies downstream of U.S. pulp mills. Process changes in pulp manufacture, including the increased use of chlorine dioxide as a bleaching agent, have markedly reduced dioxin discharges and tissue levels in fish living in mill receiving waters.

Chronic Toxicity

• is "probably attributable" to a reduction in mass loadings of non-chlorinated compounds that are indirectly measured by the bulk analytical parameter chemical oxygen demand (COD);

• "may" also reflect an incremental reduction in the potential for formation of dioxin (2378-TCDD) and furan (2378-TCDF); and

• (may also reflect) a reduction in the mass loadings of all chlorinated compounds which can be measured by the bulk analytical parameter adsorbable organic halides (AOX).

AOX as an Indicator of Polychlorinated Compounds

AET urges EPA to eliminate such scientifically flawed notions from any preamble or text in the development of final effluent guidelines.

Totally Chlorine-free Technology

Since there is no environmental justification for implementation of TCF technologies, AET urges EPA to eliminate such encouragement from any preamble or text in the development of final effluent guidelines.

In "the Notice" on page 36838, EPA makes the following statement: "The incremental environmental benefits that the Agency can attribute to the use of extended delignification (e.g., OD or extended cooking) in addition to complete (100 percent) substitution include reduced chronic toxicity to some aquatic life species. This reduced chronic toxicity is probably attributable to a reduction in mass loadings of certain non-chlorinated compounds that are indirectly measured by the bulk analytical parameter chemical oxygen demand (COD). The reduced chronic toxicity may also reflect an incremental reduction in the potential for formation of dioxin (2,3,7,8 TCDD) and furan (2,3,7,8 TCDF), which at many mills is no longer measurable by current analytical methods at the end-of-pipe, and a reduction in mass loadings of all chlorinated compounds which can be measured by the bulk analytical parameter adsorbable organic halides (AOX)." The foregoing statement is not supported by peer-reviewed science nor international scientific opinion and fundamental chemistry. The following analysis examines each of the assertions in the foregoing statement.

COD and Chronic Toxicity

COD is the oxygen equivalent of the organic matter in pulp and paper effluent samples that is oxidizable by chromic acid. Typically COD is the chemically oxidizable organic chemicals released from the wood in pulping and bleaching.

The EPA's conclusion appears to be based partly on a technical paper by Folke, 1995. This particular paper refers to a concept called "response index" to evaluate different pulp and paper mill effluent influences in model ecosystems. The "response index" is a measure of the intensity in change of the response of a parameter studied. The developers of the "response index" (Lehtinen et al., 1992) have said: "the procedure should be regarded as a semiquantitative integration of the results obtained. ... The method must be regarded as relatively subjective since the significance of the responses obtained on different biological levels is still not understood..." Folke's analysis is flawed for the following reasons:

• in the first place Folke attempts to make a relationship between a "response index" and COD. This is in spite of the fact that the "response index" "must be regarded as relatively subjective since the significance of the responses ... is still not understood. More importantly the "response index" is not a measure of "sub-lethal" toxicity and so a relationship between "chronic toxicity" and COD cannot be deduced from "response index" vs. COD.

• even assuming the "response index" is a basis for comparison, the analysis is deficient in that it compares treated and untreated effluents with and without oxygen delignification. Therefore it is impossible to distinguish the impact of extended delignification, decreased washing losses, condensate treatment and/or secondary biological treatment on the "response index." Therefore it is impossible to draw the conclusion that COD reduction afforded by extended delignification in and of itself contributes to a lower "response index."

• the correlation of the "response index" to COD is flawed in that Folke excluded a data point in the development of the correlation (without supporting argument). This data point is for TCF operation which would include extended delignification technology. Exclusion of the data point increases the correlation coefficient. Furthermore, Folke forced the "line" to go through the origin although there is no data point to support such a "forcing" condition. Not surprisingly, this further enhances the correlation coefficient as shown in Figure 1, reproduced from Folke, 1995.

• However, when the all the data points from the original reference (see Tana et al. 1996, Figure 23) are included, the relationship between "response index" and COD becomes weaker (see Figure 2) i.e., the correlation coefficient declines to 0.56. Moreover, the original regression line does not pass through the origin.

• the highest data point has a COD of 93 kg/ADt and is identified by Folke, 1995 as a mill with 5% chlorine dioxide substitution, hypochlorite bleaching, and secondary effluent treatment. It is inconceivable that a bleached kraft mill with secondary treatment would have 93 kg/ADt COD after treatment. Furthermore, in earlier publications (Lehtinen et al., 1991) this mill is identified without secondary treatment - so which is it? If this questionable data point is removed, there is no correlation between COD and "response index" as the correlation coefficient declines to 0.25, (see Figure 3).

• using the original data, the regression lines do not pass through the origin (Figures 2 and 3) indicating a "response" with "zero" COD, thus indicating confounding factors are influencing the analysis.

• finally, (assuming the "response index" is a basis for comparison), there is no significant difference in "response index" over the COD range 25 - 45 kg/ADt (the range proposed by EPA for possible BAT limitation) as shown in Figure 4.

• in the same paper, while Folke suggests COD is a useful indication of the sublethal toxicity, he also says it does not provide the full story, suggesting it is the source of the COD that is important.

• Folke has also said (Folke et al., 1993 underline added for emphasis): "the COD of a biologically treated effluent represents the fraction of the organic substances in an effluent that the natural ecosystems cannot readily degrade, but provides no indication as to whether these substances are harmful."

Martel et al. (1994) investigated the EROD-inducing potential of biologically treated mill effluents. While many effluents did not cause statistically significant EROD induction, for those that did, the potential could not be linked to a particular bleaching or pulping process. There was no significant difference between mills with and without extended delignification.

Priha, (1996) investigated the final treated effluent from 15 Finnish pulp mills using five laboratory sub-lethal (chronic) and lethal toxicity tests. Importantly, the study says (underline added for emphasis): "Comparison of the biological responses and effluent monitoring parameters shows that biological impact cannot be predicted from either the content of organic matter, measured as BOD and COD or the AOX content."

"... likewise, a high COD level did not indicate a high biological response..."

Potential for Dioxin Formation and Chronic Toxicity

This argument has fundamental flaws. In the first case EPA asserts that the potential for formation of dioxin and furan is decreased with extended delignification. This notion does not withstand scientific scrutiny.

An examination of final effluent from 17 kraft mills for dioxin and furan (Shariff et al., 1996) using the best detection techniques available, showed there was no difference in final effluent values for mills using ECF technology with or without extended delignification. Furthermore, a proposed model explaining the lack of formation of such compounds with ECF bleaching, determined that the potential for formation is a function of the ratio of atomic chlorine to the lignin content of the unbleached pulp. The potential for formation is not governed by the absolute value of the lignin content of the unbleached pulp. In other words, the potential for formation is independent of the extent of delignification of the unbleached pulp.

In the second case, in the data presented in EPA's supporting document, "Pulp and Paper Mill Data Available for BAT Limitations Development," specifically Table 3.9 p. 3-12, neither dioxin nor furan was detected at mills operating without extended delignification.

EPA notes that there are examples of ECF operation indicating that detectable concentrations of dioxin and can be generated in bleach plant effluents, (Section XI: (c) p. 36848). However, where dioxin and/or furan was detected, the mills were operating with extended delignification (Table 3-9, p. 3-12 "Pulp and Paper Mill Data Available for BAT Limitations Development"). Based on this observation, one could argue perhaps there is greater potential for dioxin formation with extended delignification. Casting that notion aside, the more likely explanation is high precursor levels as discussed in the model proposed by Shariff et al. (1996).

Shariff et al. (1996), also examined the potential ecological response from ECF mills (with or without extended delignification). The conclusion reached was: "... no adverse effects caused by PCDD/Fs in the receiving ecosystem should be expected from effluents discharged by mills using chlorine dioxide as the only oxidizing agent in the first stage of bleaching..."

AOX and Chronic Toxicity

The hypothesis that AOX loadings have a relationship with chronic toxicity has been examined extensively by the international scientific community and has been proven to be false as noted in the following peer-reviewed publications.

Solomon et al. (1993) after examining the ecological risks associated with the use of chlorine dioxide for the bleaching of pulp, independent of the extent of delignification, made the following conclusion: "... chlorinated organics from mills bleaching with chlorine dioxide, employing secondary treatment and with receiving water dilutions typical of most mills in North America, present an insignificant risk to the environment..." AxegÔrd et. al., (1993) in a review of environmental impact of bleaching effluents reached the following conclusion: "... No relationship between AOX and environmental effects has been found at the current low AOX levels in effluents from chlorine dioxide bleaching..." Tana et al., (1996) in a review of the aquatic environmental impact of pulping and bleaching operations reached the following conclusion: "Toxic responses caused by whole mill effluents from modern mills (AOX below 1.5 kg/ton pulp) are generally very low, and show no variation related to the actual AOX release. This indicates that, at least in this category of mills, the chlorinated organics in the effluents do not contribute significantly to the toxicity of the effluents..." Priha, (1996) investigated the final treated effluent from 15 Finnish pulp mills using five laboratory sub-lethal (chronic) and lethal toxicity tests. Importantly, the study says (underline added for emphasis): "Comparison of the biological responses and effluent monitoring parameters shows that biological impact cannot be predicted from either the content of organic matter, measured as BOD and COD or the AOX content." McCubbin and Folke (1995) in an examination of the significance of AOX reached the following conclusions: "... AOX discharges from modern mills cannot be correlated with environmental effect..."

"... the quantities of AOX discharged by current mills are environmentally insignificant..." O'Connor et al., (1993) examined laboratory bioassay response and AOX content of pulp and paper mill effluents and reached the following conclusion: "... The absence of a significant correlation between chronic toxicity and AOX, AOX < 1000, or EOX and the fact that unbleached and bleached kraft mill effluent exhibited similar chronic toxicities, strongly indicate that the group parameters for chlorinated organic compounds are not good predictors of bioassay response for either primary or secondary treated bleached kraft mill effluent ..." Thus the EPA's contention that "reduced chronic toxicity may also reflect a reduction in mass loadings of all chlorinated compounds which can be measured by the bulk analytical parameter adsorbable organic halides (AOX)" runs contrary to the entire body of published international scientific investigations.

EPA has made assertions in "the Notice" that: "final AOX loading is an appropriate measure of the performance of in-process and end-of-pipe technologies in reducing the chlorinated organic pollutants such as dioxin and furan found in wastewaters discharged by this industryî and ñthus, EPA expects that process changes and treatment technologies implemented to reduce AOX discharges at the end of the pipe will in turn further reduce the likelihood of the formation and discharge of these chlorinated organic pollutants." The statement that AOX is a measure of the chlorinated organic compounds in waste water is true. However, the notion that it is also a measure of the likelihood of the formation and discharge of dioxins and furans and polychlorinated phenolic compounds is not supported by peer-reviewed science nor international scientific opinion and fundamental chemistry.

Berry et al. (1991) examined the relationship between AOX formation and the formation of dioxin and furan and polychlorinated phenolic compounds. The following conclusions were reached: "... We have observed that the response of the group parameter AOX to process change is decidedly different from that of the polychlorinated organic material... In each case, AOX is essentially linearly related to the chlorine multiple, while the relationships for the representatives of the polychlorinated organic material are considerably more complex and show a comparably greater rate of reduction than AOX as the chlorine multiple is reduced..." Shariff et al. (1996) examined the potential for formation of dioxin and furan with chlorine dioxide bleaching. Their analysis of final mill effluents showed no difference in presence/absence of such compounds independent of the extent of delignification and therefore independent of the extent of AOX discharge. Furthermore, their analysis confirmed that the likelihood of formation of such compounds is not governed by the AOX discharge as asserted by EPA but by the ratio of atomic chlorine to the lignin content of the unbleached pulp (independent of the extent of delignification). This work confirms the findings and conclusions of Berry et al. (1993).

In "the Notice", EPA makes the following statements: "... Only one U.S. bleached papergrade kraft mill employs a TCF process, and it produces a market pulp of somewhat less than full market pulp brightness..."

"... European mills have achieved at or near full market brightness for limited periods..."

" ...The limited range of papergrade TCF products currently produced and sold in the U.S. market indicates that TCF is not yet available to make the full range of products produced by ECF or similar chlorine-based processes. Nonetheless, EPA continues to strongly encourage further development and implementation of TCF technologies and products..." AET is concerned that EPA is "strongly" encouraging implementation of TCF. International scientific opinion and peer reviewed science shows there is no additional environmental benefit compared to ECF-based technology. Therefore, it seems inappropriate to encourage the implementation of a technology which has no additional environmental benefit. This opinion is supported by the following investigations.

Tana et al., (1996) in a review of the aquatic environmental impact of pulping and bleaching operations reached the following conclusions: "...A comparison of the toxic responses of bleach plant and whole mill effluents from mills using different schemes for non-chlorine bleaching, i.e., ECF versus TCF bleaching, shows that neither technical concept invariably produces effluents with lower toxic potency. Thus, according to studies made to date, ECF and TCF bleaching are not possible to separate in terms of their potential to produce detrimental effects in the aquatic ecosystem..." Grahn et. al. (1995) in an examination of bleach plant and whole mill effluents from the same mill operating ECF and TCF came to the following conclusion: "... it may be concluded that no clearcut differences in biological effect-inducing potential was found between effluents from production of ECF-pulp and TCF-pulp. As a final conclusion, it may thus be stated that the overall effect picture was weak and no significant differences in the power to produce long-term toxicity at the ecosystem level was demonstrated between the effluents from the two bleaching concepts..." Priha, (1996) investigated the final treated effluent from 15 Finnish pulp mills using five laboratory sub-lethal (chronic) and lethal toxicity tests. Importantly, the study says: "... No systematic differences in biological activity between effluents treated in aerated lagoons or activated sludge plants could be detected. Likewise, the bleaching concept did not explain the variations in biological responses..." Nelson et al., (1995) investigated biotreated laboratory bleaching effluent from 4 different bleaching concepts, one ECF and three TCF and their biological response using five marine bioassays. Importantly, the study says: "... the treated ECF and TCF bleach filtrates showed similar toxicity to all species except Nitzschia, where the treated TCF effluent was more toxic..."

References

1. Axegård, P., Dahlman, O., Haglind, I., Jacobson, B., Mörck, R. and Strömberg, L. Nordic Pulp and Paper Research Journal. "Pulp Bleaching and the Environment - the Situation 1993." No. 4. p. 365. (1993).

2. Berry, R.M., Luthe, C.W., Voss, R.H., Wrist, P.E., Axegård, P., Gellerstedt, G., Lindblad, P-O. and Pöpke, I. Pulp and Paper Canada. "The Effects of Recent Changes in Bleached Softwood Kraft Mill Technology on Organochlorine Emissions: An International Perspective." 92:6 (1991).

3. Eco-System Recovery: Liftings of Fish Advisories for Dioxin Downstream of U.S. Pulp Mills: An Analysis of US EPA's Fish Consumption Database, Bans and Advisories File. Alliance for Environmental Technology, Washington D.C., August 1995.

4. Folke, J. 1995 International Non-Chlorine Bleaching Conference Proceedings, "Does COD Provide a Useful Indication of the Sub Lethal Toxicity of ECF and TCF Effluents?". Pulp & Paper, San Francisco.

5. Lehtinen, K-J., Tana, J., Karlsson, P., Engström, C., Mattsson, K., Hemming, S., Hemming, J., and Fugleberg, A-L. "Chemical Characterization and Effects in Baltic Sea Littoral Mesocosms of Treated and Untreated Effluents from Bleached Hardwood Kraft Pulp Production. Publication of the Water and Environment Administration - Series A 105. ISBN 951-47-6321-1. p. 109-125.

6. Folke, J., Lehtinen, K-J., Landner, L. and McCubbin, N. TAPPI 1993 Environmental Conference Proceedings, "Simplified Bioassays and Chemical Analyses to be Used for Regulatory Purposes in the Pulp and Paper Industry." TAPPI PRESS, Atlanta.

7. Grahn, O. and Grotell, C. Swedish Environmental Research Group. "Effects in Model Ecosystems of Bleach Plant and Whole Mill Effluents from Production of TCF-and ECF-Bleached Softwood Kraft Pulp at the Södra Cell AB Mill Värö Bruk. September, 1995.

8. Lehtinen, K-J., Axelsson, B., Kringstad, K. and Strömberg, L., Nordic Pulp and Paper Research Journal. "Characterization of Pulp Mill Effluents by the Model Ecosystem Technique. No. 2, p. 81. (1991).

9. Martel, P.H., Kovacs, T.G. and Voss, R.H. 1994 Proceedings International Fate and Effects of Pulp and Paper Mill Effluents Conference. "Effluents from Canadian Pulp and Paper Mills: A Recent Investigation of Their Potential to Induce Mixed Function Oxygenase Activity in Fish." St. Lucie Press, Florida.

10. McCubbin, N. and Folke, J. Pulp & Paper Can. "Significance of AOX vs Unchlorinated Organics." 96:2. p. T263. (1995).

11. Nelson, P., Stauber, J.L., Gunthorpe, L., Deavin, J.G., Munday, B.L., Krassoi, R. and Simon, J. 1995 Proceedings, International Non Chlorine Bleaching Conference . "Toxicity Testing of Effluents from ECF and TCF Bleaching of Eucalyptus Pulps." Pulp & Paper, San Francisco.

12. O'Connor, B.I., Kovacs, T.G., Voss, R.H., Martel, P.H. Journal of Pulp and Paper Science. "A Study of the Relationship Between Laboratory Bioassay Response and AOX Content for Pulp Mill Effluents." 19:1 (1993).

13. Priha, M. 1994 Proceedings International Fate and Effects of Pulp and Paper Mill Effluents Conference. "Ecotoxicological Impacts of Pulp Mill Effluents in Finland." St. Lucie Press, Florida.

14. Shariff, A., Ahlborg, U., Axegård, P., Rappe, C. and van Heiningen, A., "An Assessment of the Formation of 2,3,7,8-TCDD and 2,3,7,8-TCDF when Chlorine Dioxide is the Oxidizing Agent in the First Stage of Chemical Pulp Bleaching." Washington (1996).

15. Solomon, K., Bergman, H., Huggett, R., MacKague, B. and MacKay, D. "A Review and Assessment of the Ecological Risks Associated with the Use of Chlorine Dioxide for the Bleaching of Pulp". In press. Pulp & Paper Canada (1993).

16. Tana, J. and Lehtinen, K-J., "The Aquatic Environmental Impact of Pulping and Bleaching Operations - An Overview." Finnish Environmental Agency, Helsinki, (1996).