Environmental Research of the Federal Ministry of the Environment, Nature Conservation and Nuclear Safety - Water Economy Research

Report 200 24 233

Environmentally compatible cooling water treatment chemicals

by Dipl. Geogr./Hyd. Stefan Gartiser Dipl. Hyd. Elke Urich Hydrotox GmbH, Freiburg

On behalf of the Federal Environmental Agency Berlin, April 2002

I Report Cover Sheet

1. Report No. 2. 3.

UBA-FB 200 24 233

4. Report Title

Water economy

Environmentally compatible cooling water treatment chemicals

5. Autor(s) (Family Name(s), First Name(s)) 6. Report Date

Gartiser, Stefan; Urich, Elke

7. Publication Date

8. Performing Organisation(s) (Name, Address)

9. UFOPLAN-No.

Hydrotox GmbH Boetzinger Str. 29 D-79111 Freiburg

200 24 233
10. No. of Pages

106 + 91 (annex)
11. No. of References

103 + 71 (annex)
12. Sponsoring Agency (Name, Address) 13. No. of Tables

German Federal Environmental Agency Postfach 33 00 22, D-14191 Berlin (Germany)

15. Supplementary Notes

20
14. No. of Figures

11
This project was commissioned in the form of a grant on the basis of costs as partial financing to the recipients (Grant Decision Z 1.6-25106/182 of 31.01.00) In Germany about 32 billion m3/a cooling water are discharged from industrial plants and the power industry. These are conditioned in part with biocides, scaling and corrosion inhibitors. Within the research project the significance of cooling water chemicals was evaluated, identifying the chemicals from product information, calculating their loads from consumption data of more than 180 cooling plants and investigating the basic characteristic data needed for an environmental hazard assessment. Additionally, the effects of cooling water samples and products were determined in biological test systems. Batch tests (shock treatments) were performed under defined conditions in order to measure the inactivation of cooling water biocides. Generally the cooling water samples only showed low ecotoxicity, upon considering the inactivation of the biocides with time. With systematic backtracking, the genotoxicity of the cooling water from one company in the umu test could be attributed to one biocide with isothiazolinones and Bronopol as ingredients. Measurement of the inactivation of biocides, with the luminescent bacteria toxicity test, revealed a strong correlation with the inoculum concentration and enabled a better estimation of the importance of the elimination factors degradation and adsorption. An overall balance sheet of chemical loads confirmed that the principal amounts came from open recirculation cooling systems, whereas only <10% of the plants with once-through cooling water used conditioning chemicals at all. The extrapolation of consumption data for Germany gave total loads of about 4.000 t/a oxidative and 125 t/a non-oxidative biocides. Additionally inputs of about 1.500-2.200 t/a phosphonic acids, 45135 t/a molybdate and 113-216 t/a zinc were calculated. Basic substance data sheets were documented for all chemicals applied, enabling first assessments of environmental relevance using several approaches of hazard assessment.
17. Keywords

cooling water, treatment chemicals, biocide, degradation, elimination, phosphonates, phosphonic acids, polycarboxylates, polycarbonic acids, zinc, bioassay, ecotoxicity, genotoxicity, Vibrio fischeri assay, Daphnia test, umu-assay, Ames test, balance, load
18. Price 19. 20.

II

Contents
0 1 2 2.1 2.2 2.3 2.4 2.5 2.5.1 2.5.2 2.5.3 3 4 4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.1.5 4.1.6 4.1.7 4.1.8 4.1.9 4.2 4.2.1 Summary .............................................................................................. 1 Introduction.......................................................................................... 5 Current knowledge .............................................................................. 6 Fundamentals ........................................................................................ 6 Cooling water flow ................................................................................. 7 Minimum requirements for cooling water discharges in Germany ......... 9 General environmental hazards from cooling systems ........................ 10 Cooling water conditioning................................................................... 12 Dispersants and hardness stabilizers .................................................. 12 Scale inhibitors .................................................................................... 13 Biocides ............................................................................................... 14 Goals and investigative strategy ...................................................... 18 Methods .............................................................................................. 20 Laboratory investigations ..................................................................... 20 Cooling water samples ........................................................................ 20 Product investigations.......................................................................... 24 Chemical parameters........................................................................... 24 Fluorescent bacteria test according to DIN 38412-34 and Nr. 404 of the AbwV............................................................................. 24 Alga test according to DIN 38412-33 and Nr. 403 of the AbwV ........... 25 Daphnia test according to DIN 38412-30 and Nr. 402 of the AbwV ..... 25 Ames test in conformance with DIN 38415-4 ...................................... 25 umu test according to DIN 38415-3 and Nr. 410 of the AbwV ............. 26 Elimination of biocides ......................................................................... 27 Drawing up an overall balance sheet................................................... 28 Compilation of production information materials .................................. 28

III 4.2.2 4.3 5 5.1 5.2 5.2.1 5.2.2 5.2.3 5.3 5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.4.5 5.5 5.6 5.7 6 6.1 6.2 6.3 6.4 6.4.1 6.4.2 Making a balance sheet of the emissions of cooling water chemicals............................................................................................. 29 Literature and database-research........................................................ 33 Results................................................................................................ 36 Cooling water investigations ................................................................ 36 Product investigations.......................................................................... 38 Eco- and Genotoxicity.......................................................................... 38 Identifying the source of the genotoxicity in plant 6 ............................. 38 Decrease of the biocidal effect in the fluorescent bacteria test ........... 41 Evaluation of the product information sheets....................................... 52 Evaluation of the questionnnaires........................................................ 53 Open recirculation cooling systems ..................................................... 54 Once-through cooling systems ............................................................ 56 Closed circulation cooling systems ...................................................... 58 Estimation of the total loads for Germany............................................ 58 Overview and comparison ................................................................... 64 Elimination of chemicals in cooling systems and sewage plants ......... 70 Regulatory control of cooling water discharges ................................... 70 Literature and database research ........................................................ 71 Evaluation........................................................................................... 74 Composition of cooling water............................................................... 74 Emission route for cooling water chemicals ......................................... 75 Elimination behavior of cooling water biocides .................................... 75 Choice of active substances ................................................................ 77 Biocides ............................................................................................... 77 Cooling water conditioners................................................................... 82

IV 7 7.1 7.2 7.3 7.4 7.4.1 7.4.2 7.4.3 7.4.4 7.4.5 8 9 Recommendations............................................................................. 89 Energy conservation measures ........................................................... 89 Technical solutions .............................................................................. 89 Process operation ................................................................................ 91 Evaluation and selection of cooling water chemicals ........................... 93 Water risk classes, the VCI-concept for open cooling systems ........... 93 "Benchmarking"-concept ..................................................................... 94 Plant specific evaluation of cooling water chemicals ........................... 95 TEGEWA-concept for indirect dischargers .......................................... 96 Outlook ................................................................................................ 96 Sources............................................................................................... 98 Ackknowledgements ....................................................................... 105

List of Tables
Table 1: Cooling water discharge in river basins (1995) [Mio. m 3] .............................. 7 Table 2: River flow volume balance for Germany in 1992 .......................................... 7 Table 3: Ratio of used/discharged cooling water (1995) ............................................. 9 Table 4: Charateristic data for the investigated systems .......................................... 23 Table 5: Determination of the elimination of cooling water biocides ......................... 28 Table 6: Summary of the wastewater investigations ................................................. 37 Table 7: Results of the product investigations .......................................................... 39 Table 8: Source of the genotoxicity in the cooling water from plant 6 ....................... 40 Table 9: Dosing of biocides in the circulation cooling ............................................... 42 Table 10: Experimental overview of the elimination curves of cooling water biocides ........................................................................................... 43 Table 11: Elimination of BCDMH depending on the inoculum .................................. 50 Table 12: Characteristic data for the cooling systems investigated .......................... 53 Table 13: Total amounts of the investigated chemicals in open circulation cooling ...................................................................................................... 55 Table 14: Consumption data in plants with flow-through cooling .............................. 57 Table 15: Cooling water use for thermal power plants (1995 ) ................................. 60

V Table 16: Concentrations of continuously added conditioners (concentration in the circulation water in mg/l) ................................................................. 61 Table 17: Chemical usage in open circulation cooling systems in Germany ............ 63 Table 18: Total load of cooling water chemicals in the once-through cooling of the foodstuffs industry........................................................................... 64 Table 19: Comparison of the estimated consumption data for certain biocides with data from other countries (data in kg/a on a substance basis) .......... 69 Table 20: Summarized evaluation of the eotoxicity and degradability of cooling water chemicals ........................................................................................ 73

List of Figures
Figure 1: Investigative strategy ................................................................................. 19 Figure 2: Decrease of the fluorescent bacteria inhibition with isothiazolinone .......... 44 Figure 3: Decrease of the fluorescent bacteria inhibition with QAV .......................... 45 Figure 4: Decrease of the fluorescent bacteria inhibition with DBNPA (10 mg/l) ...... 46 Figure 5: Decrease of the fluorescent bacteria inhibition with DBNPA (48 mg/l) ...... 46 Figure 6: Fluorescent bacteria inhibition with glutardialdehyde (30-160 mg/l) .......... 48 Figure 7: Fluorescent bacteria inhibition with glutardialdehyde (30-1000 mg d.s./l) . 48 Figure 8: Decrease of the fluorescent bacteria inhibition with Bronopol ................... 49 Figure 9: Decrease of the fluorescent bacteria inhibition with BCDMH (4 mg/l) ....... 51 Figure 10: Decrease of the fluorescent bacteria inhibition with BCDMH (37 mg/l) ... 51 Figure 11: Proportion of biocidal active ingredients in 101 products ......................... 52

VI

Abbreviations
AbwV AMPA AOX ATMP ATV BAT BCDMH BgVV BIG BUA CAS CHEMIS COD DBNPA DTPMP DOSE ECDIN EnviChem EC50 EDTA EDTMP EQS EC GESTIS GSBL GA Wastewater Ordinance Aminomethylenephosphonic acid Adsorbable organic halogens (X = Cl, Br, I) Aminotrimethylenephosphonic acid Abwassertechnische Vereinigung e. V. Best available techniques 1-Bromo-3-chloro-5,5-dimethylhydantoin Bundesinstitut fr gesundheitlichen Verbraucherschutz und Veterinrmedizin Brandweerinformatiecentrum Gevaarlijke Stoffen Beratergremium fr umweltrelevante Schadstoffe Chemical Abstracts Chemical information system of the BgVV Chemical oxygen demand Dibromonitrilopropionamide Dieethylenetriaminepentamethylenephosphonic acid Dictionary of Substances and Their Effects Environmental Chemicals Data and Information Network Data Bank of Environmental Properties of Chemicals 50% effect concentration Ethylenediaminetetraacetate Ethylenediaminetetramethylenephosphonic acid Environmental Quality Standard European Community Gefahrstoffinformationssystem der gewerblichen Berufsgenossenschaften Gemeinsame Stoffdatenbank Bund/Lnder Lowest ineffective dilution, alga test = lowest dilution factor at which inhibition of algal biomass growth is below 20%. Lowest ineffective dilution, Ames test = lowest dilution factor at which an induction difference as compared with negative controls of <80 (TA100) resp. <20 revertants/plate (TA98) is determined. Lowest ineffective dilution, umu assay = lowest dilution factor with an induction rate below 1.5.

GEA

GEU

VII GL Lowest ineffective dilution, luminescent bacteria test = lowest dilution factor at which inhibition of luminescence is below 20% Lowest ineffective dilution, Daphnia test = lowest dilution factor at which 90% of Daphnia retain their mobility Hazardous Substances Data Bank Hydroxyethanediphosphonic acid Induction rate in the Ames and umu tests International Uniform Chemical Information Database International Union of Pure and Applied Chemistry EC-directive Integrated Pollution Prevention and Control Kommission zur Bewertung wassergefhrdender Stoffe Lnderarbeitsgemeinschaft Wasser 50% lethal concentration Molecular weight Mean water flow Nitrilotriacetate Organisation for Economic Co-operation and Development Oslo/Paris Convention for the protection of the marine environment of the Northeast Atlantic Phosphonobutanetricarbonic acid Predicted environmental concentrations Predicted no effect concentration Quarternary ammonium compounds Register of Toxic Effects of Chemical Substances Semi-continuous activated sludge test Verband der Textilhilfsmittel-, Lederhilfsmittel,- Gerbstoffund Waschrohstoff-Industrie e.V. Verband der Chemischen Industrie Technische Vereinigung der Grokraftwerksbetreiber e.V. Growth factor in the umu test Wassergefhrdungsklasse Verwaltungsvorschrift wassergefhrdender Stoffe

GD HSDB HEDP IR IUCLID IUPAC IPPC directive KBwS LAGA LC50 MW MQ NTA OECD OSPAR PBTC PEC PNEC QAV RTECS SCAS-Test TEGEWA VCI VGB WF WGK VwVwS

0 Summary
In power plants and industrial processes non-recoverable heat released during the use and conversion of energy is removed from the industrial processes by cooling systems. Water is the most important coolant medium used. In Germany about 27 billion. m3 cooling water (109 m3) are discharged per year from power plants mainly via once-through cooling systems. To this, about 5 billion m 3 from industrial plants must be added, of which about 376 million m 3 comes from plants with open recirculation cooling systems. The water consumption of open recirculation systems amounts to only 2-5% of that of open cooling systems at equal cooling capacities. Nevertheless, the water added to the system to compensate the loss of water due to evaporation or blow down ("make-up water") regularly has to be conditioned with biocides, scale inhibitors, dispersants and/or corrosion inhibitors, in order to prevent disturbances of processes by depositions (scaling), corrosion or bio-mass growth (fouling). Within the research project the input of cooling water chemicals was evaluated, identifying the chemicals from product information, calculating their loads from consumption data of more than 180 cooling plants and investigating the basic characteristic data needed for an environmental hazard assessment. Additionally, 12 water samples from 7 companies and 11 products have been evaluated in biological test systems. The elimination of eight cooling water biocides has been determined, using the luminescent bacteria assay and batch tests with defined inoculum concentrations (30-1000 mg d.s./l). Generally, the cooling water samples showed only low ecotoxicity in the algae, Daphnia and luminescent bacteria assays if the elimination time of the biocides is considered. With systematic backtracking, the genotoxicity of the cooling water from one company in the umu-assay could be attributed to one biocide with isothiazolinones and Bronopol as ingredients. No effects of the water samples have been detected with the Ames test, although several products proved to be mutagenic in the Ames test. The elimination of biocides in batch tests, as measured with the luminescent bacteria toxicity test, showed that isothiazolinones and quarternary ammonium compounds were better removed with higher inoculum concentration due

2 to their adsorption to activated sludge. In contrast, the elimination velocity for 2,2dibromo-3-nitrilopropionamide (DBNPA) increased with increasing pH. For the oxidative biocide bromochlorodimethylhydantoin (BCDMH) only a weak dependence on inoculum concentration was observed, while Bronopol showed a distinct toxicity at low inoculum concentrations even after 8 days. Therefore, the test conditions for the determination of elimination curves, which determine the period the circuit must be closed after a shock treatment with non-oxidizing biocides according to Annex 31 of the German Waste Water Ordinance, must be clearly defined. Inactivation curves performed applying the test conditions of the VCI-working group "Biocides in cooling systems" with high inoculum density (activated sludge with 500 mg d.s./l) favor elimination by adsorption, and the test design corresponds to an inherent biodegradation test. Comparable biomass concentrations normally were not found in cooling systems. If additional information is required, especially for directly discharged cooling water, results about ready bio-degradation and/or elimination curves at lower inoculum concentrations (i.e., 30 mg d.s./l corresponding to the test conditions of the OECD 301 "Ready bio-degradability" tests) should be demanded. The overall accounting of chemical loads in a balance sheet confirmed that the principal amounts came from open recirculation cooling systems, whereas only <10% of the plants with once-through cooling water used conditioning chemicals at all. The extrapolation of consumption data for Germany gave total loads of about 4000 t/a oxidative (mainly chlorine, chlorine release agents, BCDMH and hydrogen peroxide) and 125 t/a non-oxidative biocides (mainly isothiazolinones, DBNPA and quaternary ammonium compounds). Additionally, inputs of about 1500-2200 t/a

phosphonic acids, 45-135 t/a molybdate and 113-216 t/a zinc were calculated.

The questionnaire for actual cooling systems in operation uncovered distinct capacities for improvement. In 34 out of 110 facilities with open recirculation cooling systems ground water was used for cooling purposes, and in 12 cooling systems drinking water (in three systems more than 100,000 m 3 drinking water per year!). Divergent from the requirements of Annex 31, biocides were continuously added in some open recirculation and flow-through cooling systems. Among these was one

3 power plant using salt water as coolant, as well as plants of the chemical and foodstuff industries. Of course, for foodstuffs hygiene requirements (product safety) are more important than the prevention of biofouling in the cooling system. According to the operators statements, the recirculation water of closed cooling systems is often discharged indirectly via municipal treatment plants and only in isolated cases directly into the recipient water. No luminescent bacteria test results were available for two thirds of the cooling systems, although 40% of them directly discharged the cooling water. Only in some cases did operators indicate that elimination curves of the biocides used have been submitted. As a rule, only the period of time for which the circuit must be closed after a treatment with biocides is documented as specified by the producers of conditioning chemicals. Concrete examples have also been presented in which the usage of chemicals has been reduced up to 90% by simple technical or organizational measures (cleaning, shading of cooling towers from the sun). Basic substance data sheets were documented for all chemicals applied, based on extensive literature and data bank/database researches, enabling first assessments of environmental relevance using several approaches of hazard assessment. For some chemicals (e.g., butylbenzotriazole, chlorotolyltriazole, tetraalkylphosphonium chloride) considerable data gaps exist. With reference to the BREFdocument of the EU-Commission about "the application of best available techniques to industrial cooling systems", different approaches regarding the selection and optimizing of cooling water chemicals are described. There is a clear confirmation that this issue cannot be examined separately from the complex thermodynamic processes, the water quantity available and the site specific characteristics. A combination of emission- and water-quality-based criteria is recommended to assess cooling water chemicals. The advantage of emission-based approaches based on the classification system of harmful water pollutants according to the European Rphrases of the dangerous substances directive is that, along with the aquatic ecotoxicity, other protection areas such as health aspects or soil conservation are considered. Additionally, insufficient databases were considered for the assessment, following the precautionary principle. However, in order to draw attention to the loads emitted, both the consumption amount as well as the elimination in cooling systems

4 and (for indirectly discharged water) municipal treatment plants should be emphasized, as described in the TEGEWA-concept for indirect discharges. The advantage of water-quality based approaches such as the "benchmarking"-concept based on the predicted environmental and effect concentrations is that the intrinsic properties of chemicals such as bio-degradability and ecotoxicity are combined and the site specific characteristics are considered. Nevertheless, this approach focuses on the environmental quality standards for surface water derived from chemical risk assessment, and the rule of load minimizing seems to be less important when the water flow capacity of the recipient water is considered to be sufficient. The determination of toxicity loads (=effect concentration multiplied by load) is a possible further development of the "benchmarking"-concept. Prequisites for the assessment of conditioning cooling chemicals are that chemicals can be identified unambiguously in product descriptions and that data gaps will be closed.

1 Introduction
In power plants and industrial processes non-recoverable heat released during the use and conversion of energy is removed from the processes by cooling systems. Due to its high heat capacity water is the most important coolant medium used. Apart from surface water from rivers and lakes also sea, ground or drinking water is used for cooling purposes. Along with the organic and inorganic constituent compounds of this water, nonnegligible amounts of air pollutants, which might cause scaling, growth of microorganisms and corrosion, are also washed out by cooling water due to the high air turnover of cooling towers. Hence, the cooling water often is conditioned with dispersants, corrosion inhibitors and biocides. As wastewater treatment of cooling water is usually not applied, these chemicals are discharged with the blow down into the sewer or (from directly discharging plants) into the receiving water. Within the project a systematic evaluation of the input of cooling water chemicals into German surface water was carried out. To accomplish this, the chemicals used were identified from product information, their loads were calculated from consumption data of more than 180 cooling plants, and the basic characteritic data needed for an environmental hazard assessment were compiled. Additionally, the effects of cooling water samples and products were determined in biological test systems. Batch tests were performed under defined conditions in order to measure the elimination of cooling water biocides.

2 Current knowledge
2.1 Fundamentals Cooling systems can be distinguished as once-through systems, as well as open and closed recirculation systems, and their combinations (Held and Schnell 2000, Anonymous 2001c). In once-through cooling systems cooling water is used without recirculation, i.e., the warmed water is directly discharged into the receiving water. Often once-through cooling is applied in combination with a cooling tower, where the cooling water is trickled in order to remove part of the heat via evaporation cooling. Once-through cooling systems demand a large water supply. For instance, power stations with a difference between in- and outlet temperature of 10C consume, as a rule, about 3.5 m 3 cooling water per 100 MW installed electric capacity (Fichte et al. 2000). Open recirculating cooling systems are wet cooling circuits open to the air, where the water used for cooling purposes is cooled down by evaporation. As a first approximation one can assume that in open recirculating systems 70% of the heat amount is removed by evaporation. The evaporation loss depends on the cooling capacity and the climatic conditions. As a rule of thumb it can be assumed that, per 10C temperature elevation, 1.1% to 1.6% of the circulating water flow evaporates in Central Europe (Fichte et al. 2000, Sommer 1988). Additionally, droplet losses of about 0.1% of the circulating water flow are emitted. Hereby, cooling water ingredients are usually concentrated by a factor of 2-4. The concentration factor is adjusted via the blow down (draining of cooling water to the recipient water body or municipal treatment plant). The evaporation losses and the blow down are compensated by the make-up water. The fresh water supply of open recirculating systems amounts to only 2-5% of that from once-through cooling systems at equal cooling capacity. Besides the above mentioned, there are also closed circuit cooling systems (dry aircooling), which are operated without wastewater emissions and are usually applied at high process temperature levels above 50C. Hybrid cooling systems combine the wet and dry cooling principles.

7 2.2 Cooling water flow The discharge of cooling water into surface water has substantial significance for water economy. The water resource balance shows that about 40% of the precipitation drained off in Germany is used for cooling purposes, whereas significant differences in the river basins can be observed because of to regional industrial main areas and the varying water flow of the principal water recipients. For example, the proportion of cooling water from the total flow of the Elbe and Rhine Rivers is above 60%, while the proportion from the Danube River amounts only to about 10% (Table 1; Table 2). Table 1: Cooling water discharge in river basins (1995) [Mio. m 3]
River basin Cooling water discharges of mining and industry direct dischargers Donau 536,7 500,5 Rhein 3.621,7 3.527,7 Maas 17,1 15,1 2,0 n.a. Ems 34,8 33,0 1,8 Weser 175,9 156,4 19,6 Elbe 597,4 532,8 64,6 coast and sea 16,7 15,6 1,2 880,3 Oder 45,5 45,4 0,1 1,0 sum 5.045,9 4.826,4 219,4 27.347,7

indirect dischargers 36,3 93,9 Cooling water discharges from power industry for public supply 1.974,6 12.603,0
Source: Statistisches Bundesamt Fachserie 19, Reihe 2.2, 1998 Maas=sum of cooling water of the Rur, Schwalm and Niers Rivers

62,1 4.647,8 7.178,9

Table 2: River flow volume balance for Germany in 1992

coast and sea

River basin Average runoff into BRD [m /s] Average runoff from BRD [m /s] Runoff from area of BRD [m /s] Cooling water effluents in total [m3/s] Proportion of cooling water from MQ with runoff into BRD Proportion of cooling water from MQ from area of BRD
3 3 3

Donau 579 1346 767 80 6% 10%

Rhein 1225 2043 818 514 25% 63%

Maas

Ems

Weser

Elbe 253

Oder
n. a

sum 2057 4701 2644 1027 22% 39%

32 32 1

111 111 3

347 347 153

610 357 247 40%

201 201 28 14% 14%

11 11 1
n. a

2%
3

3%

44%

69%

13%

Source: Statistisches Bundesamt Fachserie 19, Umweltkonomische Gesamtrechnungen, August 1994 Statistisches Bundesamt Fachserie 19, Reihe 2.2, 1998, changed to m /s MQ = Average runoff in 1992 of the respective draining areas of rivers; BRD=Federal Republic of Germany

The power plants for public supply, with 84% of all cooling water discharges, are the most important dischargers. Only 219 million m 3 cooling water from mostly small companies were indirectly discharged via municipal sewage treatment plants, so the cooling water proportion of municipal wastewater amounts to only 2-3 % of the total by volume. Thus, nearly all the volume of cooling water (>99%) is discharged directly

8 into the receiving water course. Conequently the chemicals used for cooling water conditioning immediately enter the receiving water, so that a particular risk potential might arise, unless the chemicals are inactivated in the cooling system itself. Considering the different industrial sectors of cooling water dischargers (cf., table 3) it is evident that, next to the power plants for public supply, in particular the chemical, mining and metal industries are the principal dischargers of cooling water. From the proportion of "used" and "discharged" cooling water a "utilization factor" can be calculated. This factor gives an indication of the importance of open recirculating cooling systems in the respective industrial sector. (The data on cooling water utilization also contain the recirculated volume and multiple-shift uses for different purposes.)

9 Table 3: Ratio of used/discharged cooling water (1995)

Water used for cooling purposes Discharged Water used for cooling water Utilisation factor *) cooling without treatment *1000 m3 *1000 m3 Power plants for public supply Mining industry Foodstuff and tobacco industries Textile industry Leather industry Wood manufacturing Paper and printing industry Coking plant and petroleum processing Chemical industry Rubber ware production Glass, ceramics and stone commerce Metal products and manufacturing Engine construction Production of office machines, electrical engineering Vehicle construction Production of furniture etc., recycling Sum of industrial cooling systems Grand total of all cooling systems
Reference:

61.759.994 5.616.335 905.610 176.554 4.170 41.723 683.511 2.415.387 11.333.036 640.606 382.479 5.091.455 236.904 393.444 1.143.730 50.581 29.115.525 90.875.519

27.347.665 812.998 161.889 145.123 317 11.609 390.528 127.419 2.488.627 65.111 28.497 616.186 26.284 59.964 108.069 3.261 5.045.882 32.393.547

2,3 6,9 5,6 1,2 13,2 3,6 1,8 19,0 4,6 9,8 13,4 8,3 9,0 6,6 10,6 15,5 5,8 2,8

Statistisches Bundesamt Fachserie 19, Reihe 2.2, 1998

*) The Utilisation factor here refers to the discharged cooling water and not to the make-up water!

2.3 Minimum requirements for cooling water discharges in Germany In Germany the discharge of cooling water is regulated in Annex 31 of the Framework regulation for wastewater. Here inter alia the following requirements are given: With the exception of phosphonates and polycarboxylates exclusively complexing agents which are readily bio-degradable may be used, The wastewater must not contain chromium, mercury or organometallic compounds, The concentrations for chlorine, AOX, COD, phosphorus and zinc are limited,

10 For fresh water cooling systems a shock treatment with microbicidal substances is limited to oxidative biocides (chlorine, chlorine dioxide, hydrogen peroxide, ozone). After a shock treatment with a biocidal substance the blow down of recirculating cooling systems is only allowed if the luminescent bacteria toxicity does not exceed GL =12 (GL= Lowest ineffective dilution factor, LID). Annex 31 currently is being revised and will be in force in 2002 (Anonymous 2001). A background paper (draft of the Bund/Lnder GK 21/41 from 17.12.2001) will also be published in 2002. In the course of the implementation of the EC-directive 96/61/EC concerning Integrated Pollution Prevention and Control (IPPC-directive) an extensive "Reference Document on the Application of Best Available Techniques to Industrial Cooling Systems" was elaborated, which is available in the internet (http://eippcb.jrc.es). The aim of the IPPC-directive is to optimize the operation of industrial plants, so that while considering energy efficiency and waste avoidance no substantial pollution of the environment will be generated. Hereby measures for the improvement of one environmental compartment (e.g., water) shall not lead to additional stress of another compartment (e.g., air). The reference document offers a comprehensive documentation for the selection of cooling systems, technical descriptions and potential environmental effects. It is clear that the cooling system cannot be considered separately from the industrial process and location. By optimizing the overall process substantial amounts of energy can often be saved. In addition, the excess energy should be used insofar as possible, for example for hydrothermal heating projects. Although in the reference document approaches for evaluating the chemical additives in cooling systems are described (cf., sect. 7.4), until now there has been no systematic presentation of the basic data required for this, both on the input side (consumption data), as well as on the material, chemical side (degradability, ecotoxicity, genotoxicity, bioaccumulation). 2.4 General environmental hazards from cooling systems In the operation of cooling systems a complex field of tensions between various usage interests and environmental conflicts arises. The water consumption is highest

11 for once-through cooling systems and for larger power plants this can exceed several m3/s. Depending on the mesh width of the inflow rates to the cooling system and the flow rate, substantial numbers of fish, especially young ones, can be sucked in and killed (up to 25 fish per 1000 m 3; Anonymous 2000). The temperature increase in the surface water leads to a reduction of the oxygen solubility in the water together with an increase in the metabolic activity. Since this can lead to a shift in the species spectrum in the waters (LAWA 1991), heat load plans have been prepared for the waters. The EU-Guideline 78/659/EWG specifies for Salmonid and Cyprinid waters, among other things, the maximal permissible temperature elevations (1.5C and 3C) and maximal temperatures (21.5C and 28C, and, during the spawning period of cold-water fish, for certain waters 10C; 78/659/EWG 1978). For cooling towers a large part of the heat burden is released as latent heat (evaporation) and causes an increase of the air temperature, which can lead to changes in the local microclimate (VDI 3784: 1986). For large power plants naturaldraft wet cooling towers are used, for which the construction height provides sufficient force to drive the air current. For ventilator cooling towers the necessary amounts of air are, in contrast, introduced by forced air blowers, for which electrical energy must be provided (corresponding to 0.5-2% of the amount of emitted heat energy, Anonymous 2000). For open recirculation cooling systems the water consumption is usually reduced by ca. 95%-98% compared to once-through cooling systems at equal cooling capacity. At the same time, however, the electrical energy needed for the pumps is increased by ca. 50% thus amounting to ca. 1.5% of the amount of emitted heat energy. The evaporation losses can be taken to be ca. 0.40.7 l/s per 1000 MW of output electricity (Wunderlich 1978b). This leads to an increase in the concentration of the constituent compounds in the water, so that often a purification of the water and/or a conditioning of the cooling water is necessary, which then requires the addition of chemicals to the receiving waters. The most urgent goal of the plant management of cooling systems, however, is their efficiency and protecting the system against depositions (scaling), corrosion and biomass growth (fouling). The formation of depositions on the cool water side of a heat exchanger or pipeline interferes with heat transfer and increases the loss of pressure, so that the performance is substantially reduced. Ultimately, this leads to a

12 higher water consumption and must be compensated by an increased application of energy. Thus, a calcium deposit 0.5 mm in thickness reduces heat transfer in condensers by ca. 20% (Todutza and Steinlein 1990). Corrosion processes not only damage the system, but also increase the risk of leaks on the production side. In addition, the corrosion products endanger the waters. A control of the biomass growth is performed with the additional goal of minimizing the microbiological risks arising from the cooling plant. It is known, for example, that thermophilic human pathogens, especially Legionella pneumophilia, which causes a severe pneumonia (Legionnaires disease), can be found in cooling systems (States et al. 1987, Kusnetsov et al. 1997, Werner and Pietsch 1991, Howland and Pope 1983, Kusnetsov et al. 1993). Guidelines for controlling Legionella in cooling systems are available (Anonymous 2001a). 2.5 Cooling water conditioning For the prevention of scaling in recirculation cooling systems, dispersants and hardness stabilizers are added. In addition, corrosion inhibitors and biocides are used, whereby there are overlaps between the individual groups (e.g., phosphates act both as hardness stabilizers and corrosion inhibitors). 2.5.1 Dispersants and hardness stabilizers The precipitation of salts due to their exceeding their solubility limits is termed scaling. Of particular interest in cooling systems is the precipitation of calcium carbonate and calcium phosphate, and to a limited extent also calcium sulfate and silicates. The hardness of the water can also be reduced by active decalcification (precipitation with calcium hydroxide). The residual hardness is either removed by conversion of the carbonate hardness into non-carbonate with acids (primarily hydrochloric and sulfuric acid) or stabilized through the addition of hardness stabilizers such as orthophosphate, polyphosphates and phosphonic acids. The ready hydrolysis of polyphosphates to orthophosphate and the associated danger of calcium phosphate deposition led to the development of stable phosphonic acids, which are added in sub-stoichiometric amounts (Andres et al. 1980). The most important phosphonic acids used in the field of cooling water treatment are aminotrimethylenephosphonic acid (ATMP), hydroxyethanediphosphonic acid (HEDP) and

13 phosphonobutanetricarbonic acid (PBTC). Organic polymers based on polyacrylic acid, polymetacrylic acid, polymaleic acid and polyacrylamide (so-called

polycarboxylates) also have a certain hardness stabilizing effect and are often used in combination with phosphonic acids. The calcium carbonate hardness can be stabilized with, e.g., carboxymethylcellulose. To a limited extent complex formers (NTA) are used; EDTA is however excluded de facto from use because of its poor degradability. Further depositions can also be caused by the precipitation of suspended organic and inorganic particles and iron oxides. To prevent this, dispersants based on the above mentioned polycarboxylates as well as low-molecular weight anionic acids (e.g., succinates) are added. These are to be distinguished from natural products, such as lignins and tannins, and from synthetic polymers of the polyacrylic, polymetacrylic, and polymaleic acids as well as sulfonates. The transition between the hardness stabilizers and the dispersants is not clear-cut. 2.5.2 Scale inhibitors The corrosion of metals is enhanced by the presence of oxygen, salt content (especially chlorides) and a low pH, but also by deposits. During oxygen corrosion metal ions are dissolved at the metal surface, which acts as the anode, while in the cathodic reaction oxygen is reduced to hydroxide ion and a high pH is produced locally (Anonymous 1991). Of particular importance is microbially induced corrosion, which is caused by acidic metabolic products as well as the anoxic/anaerobic conditions within biofilms. Sulfate-reducing bacteria of the genus Desulfovibrio act corrosively, by reducing the sulfate while forming hydrogen sulfide. These bacteria are among the most important in cooling systems (Koppensteiner 1973). However, corrosion can also be induced by sulfur bacteria (Thiobacillus), iron bacteria (Ferrobacillus, Gallionella) and nitrifying bacteria (Nitrosomonas, Nitrobacter). Passive (anodal) corrosion inhibitors, such as phosphates, phosphonates, nitrite, silicates and molybdates form a passive protective layer on the metal surface. The use of chromate is no longer permitted. In contrast, cathodic inhibitors like zinc or calcium carbonate, and to a limited extent also orthophosphate, form insoluble deposits which protect the metal surface by reacting with the corrosive hydroxyl ions.

14 Especially for copper and copper alloys, 1,2,3-triazoles are used as inhibitors. Mercaptobenzthiazoles may no longer be used for this purpose according to Annex 31 of the AbwV regulations. 2.5.3 Biocides The mean temperature in water cooling systems is ca. 35C and thus lies just below the temperature optimum of most microorganisms (Mattila-Sandholm and Wirtanen 1992). Biocides are used to control biologically induced deposits and corrosion processes. For cooling water systems algicides, fungicides and molluscicides are relevant. 2.5.3.1 Biology in cooling systems The growth of autotrophic algae is dependent on the presence of mineral nutrients, carbon dioxide and light energy, while the growth of heterotrophic bacteria requires organic material, which is composed of dead algae and/or the existing burden of the water or air. In principle, bio-degradable conditioning agents can also function as a carbon source. Many bacteria secrete a highly hydrated slime consisting of polysaccharides, which leads to the formation of biofilms on surfaces (biofouling). Biofilms decrease heat exchange, promote corrosion and hinder control by means of biocides. Protozoa such as Ciliates or Ambae colonize affected cooling towers as consumers, as do higher organisms such as mussels and snails, which can lead to serious disturbances. In once-through cooling systems, because of the short retention time and the requirement for a rapid elimination, fast-acting oxidative biocides are used; and in open cooling systems, non-oxidative, more stable organic biocides are called for. 2.5.3.2 Oxidative biocides The most commonly used oxidative biocide, owing to its effectiveness and low price, is chlorine or cholorine bleach (sodium hypochlorite). At the pH-values of > 8 typical for cooling system circulation, there is a reduction of the biocidal effect of the active substance, hypochlorous acid (HOCl), while hypobromous acid is still effective at pH 9. Hypobromous acid is generally generated on site by adding sodium bromide to sodium hypochlorite (NaOCl). The use of free halogens as biocides may, depending

15 on the water composition (e.g., DOC- and ammonium concentration), pH-value and contact time, lead to the formation of disinfectant by-products such as trihalomethanes, chloro- and bromoamines as well as absorbable organic halogen compounds (AOX). In the purification of drinking water chlorine is replaced in part by chlorine dioxide, in order to minimize the formation of AOX, especially halogen methanes. Chlorine dioxide reacts noticably more weakly with complex organic molecules and ammonium, consequently forming less AOX. Chlorine dioxide is also occasionally used in the cooling water field, whereby it is usually generated on location through the reaction of chlorine gas with sodium chlorite (NaClO 2). Organic chlorine and bromine release agents are used especially in open recirculation cooling systems. Here, above all, the rapidly hydrolyzing biocide 1-bromo-3-chloro-5,5-

dimethylhydantoin (BCDMH) should be mentioned. Related compounds like 1,3dichloro-5,5-dimethylhydantoin or 1,3-dichloro-5-ethyl-5-methylhydantoin are also occasionally used. Ozone is a highly effective oxidatively acting biocide. Usually, ozone is continuously added to the cooling water in very low concentrations of 0.1 to 0.3 mg/l (WaselNielen and Baresel 1997, Viera et al. 1999). Production is achieved directly on location using high voltage, In comparison with the other oxidative biocides, hydrogen peroxide is only effective at higher concentrations (> 15 mg/l; cf., van Donk and Jenner 1996) and has a short half-life. Rarely, peracetic acid is also used as an organic oxygen release agent in cooling systems. Under unfavorable conditions, peracetic acid is corrosive. This chemical is readily bio-degradable. 2.5.3.3 Non-oxidative biocides Non-oxidative biocides are used nearly exclusively in open recirculation cooling systems, where the contact time of the cooling water with the biocide suffices for a satisfactory effect. As a rule, here the biocide is added batchwise in a shock treatment. One of the most important non-oxidative cooling water biocides, a mixture consisting of a chemical belonging to the isothiazolinone family, 5-chlorine-2-methyl-4isothiazolin-3-one, together with 2-methyl-4-isothiazolin-3-one, is already effective at

16 concentrations below 1 mg/l. Isothiazolinones hydrolyze slowly (t 1/2 = 7 d at 30C and pH 8) and are not readily bio-degradable. Quarternary ammonium compounds (QAV) act through their binding to the cell membrane and are also not readily biodegradable. During passage through the sewage treatment plant they are largely eliminated by adsorption on the activated sludge. The most important representative in the cooling water area is alkyldimethylbenzylammonium chloride. The addition of dibromonitrilopropionamide (DBNPA) is also widespread in the treatment of cooling water. This compound hydrolyzes rapidly to the still partially biocidally active compounds dibromoacetonitrile, dibromoacetamide, monobromonitrilopropionamide and cyanoacetamide. Further members of the organic bromine compound group include 2-bromo-2-nitropropan-1,3-diol (Bronopol) and beta-bromobeta-nitrostyrene. Glutardialdehyde is also rather frequently used in the cooling water field. The mechanism of action is based on the denaturation of proteins. Glutardialdehyde is less toxic for aquatic life forms as compared to the other biocides, and the concentration added is correspondingly higher. This compound is readily biodegradable. Specifically for the control of algal growth additional biocides are used, such as copper sulfate, as well as photosynthesis inhibitors based on triazinederivatives. To reduce the risk of the appearance of microorganisms resistant to the added biocides combination products containing several biocides are used. 2.5.3.4 Elimination of the biocidal effects A basic requirement for cooling water biocides is that their damaging action or biocidal effects must diminish in a relatively short time, since otherwise there might be toxic effects on the surface waters, especially after the direct discharge of cooling water. This calls for a rapid hydrolysis and/or biological degradability of the biocides. For indirect emissions via municipal sewage treatment plants it has to be proven that the biological wastewater treatment is not inhibited and that the biocides are retained in the treatment plant. Preferably, the biocides should be biologically degraded. While elimination through adsorption on the activated sludge (cf., QAV) protects the

17 receiving water, this merely shifts the problem, when the collected sludge is spread on the land for agricultural or forestry use. The elimination of the biocidal effects can be assayed either in the laboratory or on site at the actual treatment plant. For the completion of such so-called elimination curves in the laboratory there are, however, no generally acceptable specifications to date. Here the manufacturers have proposed static experiments with relatively high concentrations of activated sludge (0.5 g d.s./l), in order to simulate the influence of a hypothetical biofilm in the cooling system circulation (Scheidel et al. 1996). Other authors, on the other hand, determine an elimination curve without adding any inoculum (Gartiser and Scharmann 1993, Gellert and Stommel 1995, Baltus et al. 1999).

18

3 Goals and investigative strategy

In accordance with the project description, the following goals have been set: Estimation of the emission of cooling water conditioners in flowing surface waters of the Federal Republic of Germany Determination of the introduced cooling water chemicals and investigation with respect to their ecotoxicity, genotoxicity, bioaccumulation and degradability Determination of current practices of the governmental control agencies in the individual Bundeslnder Extension of the data status on ecotoxicity, genotoxicity and biological degradability of the cooling water chemicals in use through measurements of our own Develop suggestions/proposals for the reduction and optimization of the addition of cooling water chemicals Develop a recommendation for the selection of cooling water chemicals based on the present technical state of the art The investigative strategy is based on three pillars (see Fig. 1): Literature and database research on the active ingredients/substances of the standard commercially used cooling water conditioners Drawing up of an overall accounting balance sheet of the loads and concentrations of cooling water conditioners in treatment systems Direct testing of cooling water samples, products and active ingredients with respect to their ecotoxicity and genotoxicity, as well as determining the rate of elimination of biocides (elimination curves)

19

Research
- Products - Active substances - Elimination of biocides - Ecotoxicity, Biodegradability Genotoxicity

Overall Balance
- Consumption data of plants - Wastewater concentration - Total volume loads in BRD - Total loads of active Subst.

Practice of regulatory control

Comparision of data

State of the art

Elimination of biocides/ Products

Active substances/ Products

Cooling water

Tests

Figure 1: Investigative strategy

20

4 Methods
4.1 Laboratory investigations 4.1.1 Cooling water samples Cooling water samples from 7 treatment plants in Southern and Northern Baden were investigated. All these plants have open recirculation cooling systems. Collection of the samples was as qualified test samples direct from the investigated cooling vessel or from the return flow of the circulation (DIN 38402 1991). The descriptive data of the investigated cooling water with respect to water consumption and the products added are presented in Table 4. a) Plant 1, Electroindustry The operation of a semi-conductor manufacturer has six cooling towers with a total cooling capacity of 2.5-3 MW. The cooling tower investigated has a cooling capacity of 1.2 MW and removes the heat produced by a refrigeration system. As a special feature, the concentrate of the water treatment (reverse-osmosis system with a capacity of 700 S/cm) is used as cooling water. For hardness stabilization and corrosion inhibition a product based on sodium phosphonates, sodium molybdate, sodium polycarboxylates and triazoles is added continuously. As a biocide, isothiazolinone is added in summer as needed. Through regular mechanical cleaning of the cooling vessel and an adequate shading of the cooling tower the amount of this chemical added was reduced by more than 90% compared to the previous years. The yearly consumption of isothiazolinones in the year 2000 was ca. 0.4 kg/a of active substance. At the time of sampling no biocides were being added. b) Plant 2, Plastics manufacturing industry The company manufactures molded plastic parts for the automobile industry and has several cooling towers with a total capacity of 6700 kW, which are fed with ground water. As biocide a quarternary ammonium compound is added batchwise as needed and then the outflow is closed for the next three days. The time of addition is decided upon by visual examination of the algal growth. In addition, corrosion

21 inhibitors and hardness stabilizers based on phosphonic acids, zinc chloride and dispersants are continuously added. The first sample collection on 19.07.00 took place two months after a shock treatment with biocides, and the second and third samples were taken either directly after the biocide was added and after an elimination time of three days resp. Besides the cooling water, the wastewater produced from the released steam, to which hydrazine was added, was also examined. c) Plant 3, Plastics manufacturing industry The plant manufactures PVC-foils and has a cooling tower with a capacity of 2.3-9.2 MW. Ca. 250 m3 of completely desalted cooling water are added weekly. In the nonshaded cooling tower problems with algae arise. In this event, a "heterocyclic sulfur/nitrogen-compound" (Isothiazolinone) is added batchwise in a shock treatment (total load 0.8 kg/a active ingredient). d) Plant 4, Plastics industry The plant of a manufacturer of adhesive foils has four cooling towers, whose function is to thermally reclaim solvents from activated charcoal filters. About four times a year a preparation based on isothiazolinones is added batchwise (ca. 30 liters of product/a). After a retention time of 24 h according to statements from the operator, the GL-value is ascertained to be below 12. As hardness stabilizers polycarboxylates and phosphonocarboxylates are added continuously. At the time of the collection of the first sample, there was no addition of biocide, and the second sample was taken 24 h after a biocide treatment just after the outflow was reopened. e) Plant 5, Plastics manufacturing industry The plant manufactures foamed polystyrene- and polypropylene-packaging and has three cooling towers. For the foaming of the plastics 10-12 t/h of desalted boiler feeder water (steam) is needed. The boiler feeder water is treated with a corrective material and the condensed steam (condensate) enters the open cooling system during the production process. Thereby solid materials from the production process are also carried over and are removed from the cooling water circulation with bag filters (towel filters). These filters are cleaned daily. Twice weekly 60 liters of a

22 biocidal product based on hydrogen peroxide and quarternary ammonium compounds are added. The cooling water sample was taken in the filter outflow 7 h after a batch treatment. f) Plant 6, Chemical industry The plant has two cooling water systems, which were both sampled. The plant cicrculation (KW1) has four forced-aeration cooling towers and is operated with a low compression ratio of 1.1. The second circulation (KW2) handles the central cooling system and is run at a high cycle of concentration of 3.0. Both circulations are treated continuously with the biocide 1-bromo-3-chloro-5,5-dimethylhydantoin. As needed, a product based on 2-bromo-2-nitropropan-1,3-diol and isothiazolinones is added. For corrosion inhibition phosphonic acids and tolyltriazole are continuously added. g) Plant 7, Chemical industry This pharmaceuticals producing operation set up a new cooling system for the expansion of the refrigeration plant. A portion of the drinking water needed for feeding this system is completely desalted by an ion-exchanger. The cooling water is treated twice weekly for several h with the biocide 1-bromo-3-chloro-5,5dimethylhydantoin. To inhibit corrosion phosphonic acids and triazoles are added continuously. The sample was collected 24 h after the last biocide treatment after opening the outflow.

23
Table 4: Charateristic data for the investigated systems
Firm Water consumption Principle Products Active ingredient 1 Branch [kW] [m /a ]
3

Active ingredient 2 Co nc. [%] Load [kg] Conc. [%] Loa d [kg]

Active ingredient 3 Conc. [%] Load [kg]

Active ingredient 4

[m / (a*KW)]

Concen tration ratio

Products

Bioci deadditi on

Product [kg/a]

Dis cha rge

Nr. 1 2.500 14.402 Electroindustry Nr. 2 6.700 1.761 Plastics manufacturing

5,8

RO TW G

OK

0,3

OK 1,8

Biocide 1 Product Product Biocide 3 Product Product Biocide 4 PO43Product Biocide 2 Biocide 5 NaOH Product

S S

10 457 90 660 2250 4145 40 24 900

KA

KA

CMI Triazoles HCl QAV HEDP HCl CMI Orthophosphate Polycarbonates CMI/MI H2O2 NaOH Polyethoxylate NaOH

2,4 2,5 21,0 15,1 30,0 0,9 100,0

0,2 11,4 18,9 339,8 1243,4 0,4 24,0

MI NaOH

2,4 2,5

0,2 11,4

Phosphonate

Na.molybdate (without conc.) 10,1 9,9 227,3 410,3 Zinc chloride (without conc.)

ATMP HEDP MI

20,1 9,9 0,9

452,3

Polyoxycarbonacid 1,4-Butindiol

Nr. 3 3.000 13.000 Plastics manufacturing Nr. 4 18.710 23.579

4,3

VE

OK ? OK 3,0 OK

KA

0,4

1,3

Phosphoncarbonate QAV 25,0 780,0 Non-ion. Tenside Tolyltriazole

Foil production Nr. 5 4.000 2.600 Plastics manufacturing

TW

S S

30 6240 3120 1200 7020

KA

Nr. 6 Chemical Industry Circulation 6.300 1 720.000 114, G 3 OK/1,1 Biocide 6 Biocide 7 NaOCl Circulation 43.500 2 120.000 G OK/3,0 Product Product K S

3,5

245,7

3,0

210,6

Phosphonic acid

7,5

526,5

1600 644 125 2000 7350 1000

Phosphonic acids O and BCDMH KA Bronopol Sodiumhypochlorite NaOH

25,1 9,0 13,0 3,5

161,6 11,3 260,0 257,3

MI

1,8

2,3

CMI

1,8

2,3

Tolyltriazole

3,0

220,5

Phosphonic acid

7,5

551,3

Phosphonic acids Biocide 6 K 2323 KA BCDMH 25,1 583,1 Biocide 7 S 25 Bronopol 9,0 2,3 MI 1,8 0,5 CMI 1,8 Nr. 7 6.000 58.000 9,7 TW OK Product Planning KA Polycarbonic 17,5 Phosphonic 6,3 Triazole 2,5 acid acid Chemical Industry max. 4.0 Biocide 8 S stage BCDMH 75,0 Water consumption: G=Ground water; RO=Reverse osmosis; VE=completely desalted water; TW=Drinking water; Biocide added: K= continuous; S=Batch treatment; Source: O=surface water; KA=municipal water purification plant; ingredients: CMI: 5-chloro-2-methyl-2H-isothiazolin-3-one; MI: 2-methyl-2H-isothiazolin-3-one; BCDMH: 1-bromo-3-chloro-5,5-dimethylhydantoin; Bronopol: 2-bromo-2-nitropropan-1,3-diol; HEDP: 1Hydroxyethan-1,1-diphosphonic acid; ATMP: Aminotrimethylphosphonic acid

24

4.1.2 Product investigations Altogether five biocides and one corrosion inhibitor were tested. All of these active substances are used in the plants investigated. They were tested with respect to their mutagenicity, genotoxicity and ecotoxicity. At a later time four more products as well as the biocide Bronopol were investigated in the umu test, in order to to determine the source of the genotoxicity in the wastewater of one plant by systematic "backtracking". An additional emphasis was the determination of the elimination behavior of the most important biocides applied in the cooling water field, using the fluorescent bacteria test (cf., sect. 4.1.9). 4.1.3 Chemical parameters pH-value: pH 196 pH-meter from WTW GmbH in Weilheim. Conductivity: Measuring instrument pH-LF 3001 from Neukumelektronik GmbH in Straubenhardt. CSB: Round cuvette test (Dr. Lange Co.): Two-hour oxidation with potassium dichromate, sulfuric acid, silver- and mercury sulfate at 148 C in conformance with DIN 38409 H41. Chlorine (free and total): Round cuvette test (Dr. Lange Co.): reaction with diphenylp-phenylendiamine (DPD) witih the formation of a colored substance; total chlorine determined after addition of potassium iodide. 4.1.4 Fluorescent bacteria test according to DIN 38412-34 and Nr. 404 of the AbwV The toxicity of wastewater contaminants is detected on the marine bacteria of the species Vibrio fischeri, which show a natural light production (bioluminescence) that is closely coupled with their metabolic activity. The decrease of the light intensity provides a quantitative measure of the toxic effect on the bacteria. The test is performed with the LUMIS-tox system of the company Dr. Lange, Dsseldorf. The lyophilized bacteria of the strain Vibrio fischeri NRRL-B-11177 were obtained from the same company (LCK 482). The wastewater samples were tested without further pre-treatment after salinizing with sufficient sodium chloride to give a 2% solution

25 and adjusting the pH-value to 7.0 +/- 0.2. The test result is given as the least stepwise dilution (GL-value), for which the light emission is inhibited less than 20%. 4.1.5 Alga test according to DIN 38412-33 and Nr. 403 of the AbwV The chronic inhibitory effect of the cooling water samples on the growth of Scenedesmus subspicatus, a planktonic fresh-water alga, was determined. For this purpose, a dilution series of the cooling water sample was made, without any further preparation, but adding an algal nutrient solution inoculated with a defined algae suspension (corresponding to 10 4 cells/ml) and incubating under defined light and temperature conditions. After 72 h, the number of cells was determined microscopically as a measure for the biomass. The result given is the least dilution step (GA-value), after which the measured inhibitory effect on biomass production is less than 20%. 4.1.6 Daphnia test according to DIN 38412-30 and Nr. 402 of the AbwV The acute toxic effect of wastewater on Daphnia magna STRAUS (Crustacea, clone 5 of the German Federal Health Agency) was determined. The value measured is the dilution factor G D beyond which no acute toxicity for Daphnia is detected within 24 h. The GD-value corresponds to the least dilution factor by which a wastewater sample must be diluted in order for 90% of the Daphnia to maintain their ability to swim. The pH-value of the sample was adjusted with hydrochloric acid or sodium hydroxide solution to 7.0 7.5. No other pre-treatment was performed. 4.1.7 Ames test in conformance with DIN 38415-4 The Ames test is a bacterial mutagenicity test with Salmonella typhimurium. The Salmonella-bacterial strains used are deficient mutants, which are unable to grow in histidine-free medium. These histidine-requiring mutants can back-mutate (reversion) and then are able to form colonies on minimal-agar plates. Each of the Salmonellastrains has a specific spontaneous mutation rate. The number of back-mutated bacteria (revertants) above this level provides a measure of the mutagenic potential of a substance or a sample. Certain mutagens in higher organisms are first activated by being metabolized (promutagens) or are thereby inactivated. Therefore, to the bacterial system the needed enzymes are added in the form of rat liver extract S9

26 (Moltox Co.). The test version used is based on a simplified version of the OECDGuideline 471 with the test strains TA98 and TA100. The strain TA98 detects frameshift mutagens; strain TA100 in contrast is for base pair substitution mutagens (point mutations). The cooling water samples were sterilized over a membrane filter (0.45 m). Up to 1 ml of cooling water per Petri dish could be added. Because of the substantial effort involved, the samples were initially investigated in the ScreeningTest at only one test concentration. A sample is then classified as mutagenic according to DIN 38415-4 if in one of the strains with or without S9 an induction difference compared to the control (solvent alone) of 80 (for TA100) or 20 revertants (for TA98) is induced and a dose-effect relationship is found. The G EA-value corresponds to the last dilution step at which the induction difference established for that strain is not exceeded. Since the wastewater sample in the test is diluted by a factor of 3 with medium/inoculum, the lowest possible G EA-value = 3 (nonmutagenic). The number of revertants of the negative controls should be: for TA100 in the range of 80-180 and for TA98 in the range of 15-40 revertants per plate. In testing substances or products a sample was evaluated as being mutagenic in accordance with the relevant OECD-guideline if the induction rate (ratio of the number of revertants in the test plates to the negative controls) exceeded a factor of 2 and a dose-effect relationship existed. 4.1.8 umu test according to DIN 38415-3 and Nr. 410 of the AbwV The umu test is a genotoxicity test with the gene-technologically modified bacterium Salmonella typhimurium strain TA1535/pSK 1002. The bacteria are exposed to various concentrations of the cooling water. Here gene toxins induce the so-called umuC-gene, which belongs to the SOS-repair system of the cell and which acts to prevent damage to bacterial genetic material. Through the coupling of the umuCgene promoter with the lacZ-gene for -galactosidase the activation of the umuCgene can be indirectly measured spectrophotometrically at 420 nm through the formation of a colored product from the -galactosidase substrate o-nitrophenylgalactopyranoside (ONPG). The induction rate (IR) corresponds to the increase of the extinction at 420 nm relative to the negative control. In calculating the induction rates one must take into account the growth factor, which is determined turbidometrically from the optical density at 600 nm. An inhibition of bacterial growth

27 is expressed as a reduced growth factor ("Wachstumsfaktor" or WF) compared to the controls. For growth factors below 0.5 (50% growth inhibition) the results are not evaluated. The result given is the smallest dilution step G (G EU-value), at which an induction rate < 1.5 is measured. If a different induction rate is seen upon addition of S9, the higher of the two values is taken (=G EU-value). 4.1.9 Elimination of biocides Until now there have been no generally accepted specifications for a procedure to determine elimination curves in the laboratory. On the part of the producing companies there have been proposals for static experiments with a relatively high concentration of activated sludge (0.5 g d.s./l), in order to simulate the influence of a hypothetical biofilm in the cooling circulation (Scheidel et al. 1996). However, at an UBA-Workshop on the present project the consensus was arrived at that such high biomass concentrations in cooling circulation are not usual (cf., table 5, (Gartiser and Urich 2001). In order to determine the effect of the inoculum concentration on the elimination behavior of biocides, various inocula were introduced. As a test-system the ZahnWellens test according to DIN EN 29888 or Nr. 408 of the AbwV was correspondingly adapted. The tests with activated sludge were supplemented with an inorganic nutrient solution according to DIN EN 29888; all tests were continuously stirred and aerated with an aquarium pump.

28 Table 5: Determination of the elimination of cooling water biocides Inoculum Activated sludge " " " Outflow finalclarifier Tap water Concentration 1 g d.s./l 0.5 g d.s./l 0.2 g d.s.//l 0.03 g d.s./l Comments Upper conc. Zahn-Wellens test Proposal by VCI AG "Microbiocides in Annex 31" Lower conc. Zahn-Wellens test OECD 301 A, B, C and F "ready bio-degradation" Model for microbiologically active inoculum with low d.s.-content predominantly abiotic hydrolysis

The starting concentrations of the biocides were selected on the basis of various information provided on the effective concentrations of the active ingredients in the cooling water (Baltus and Berbee 1996; Anonymous 1994, Fielden and Iddon 1997) and in part reduced further according to updated information from the manufacturers (Klautke 2001) (Table 9). As the end point, after filtration through a folded paper filter a bacterial fluorescence toxicity test was performed at dilution step 12 (based on Annex 31 to the AbwV). For low toxicity, dilution step 2 was also tested.

4.2 Drawing up an overall balance sheet 4.2.1 Compilation of production information materials Letters were sent to a total of more than 100 firms in the chemical industry that also offer product groups used in the cooling water field (including algicides, antifouling agents, bactericides, corrosion inhibitors, dispersants, biocides, inhibitors, and water chemicals). Addresses were obtained in some cases from the relevant handbooks but also to a large extent through information provided by the operators of the cooling plants. Altogether, 49 firms replied that they were not engaged in the cooling water field. Product information was sent by 22 firms. These materials were of

29 varying quality (from safety data sheets with little useful information to detailed product descriptions with ecological evaluations). In order to learn more about the substances in the products investigated in the context of preparing the balance (cf., sect. 4.2.2) the operators of ca. 50 cooling plants were requested to provide the corresponding safety data sheets. In this way information on 418 products from 35 manufacturers was obtained and evaluated. The active substances documented in these materials served as the basis for our literature- and database-research (cf., sect. 4.3). 4.2.2 Making a balance sheet of the emissions of cooling water chemicals The yearly emissions of cooling-water chemicals by the firms considered in the cooling water sampling were determined on site. In order to extend the data base a questionnaire was prepared for the operators of the cooling plants, including questions about the cooling system used, the cooling capacity, the source of the water, the annual consumption of cooling water chemicals (on a product basis), the mode of addition of biocides and the parameters controlled. Initially, this was sent to all business-controlling governmental agencies in Baden-Wrttemberg and then to the environmental agencies in all the Bundeslnder (usually to the Environmental Ministry). After the questionnaires had been passed on to the local sub-authorities and/or the operators of the cooling plants, they were then returned to us either directly or through the authorities. In some cases, Hydrotox was also provided by the authorities with lists of addresses of operators of cooling plants, and we then contacted them directly. Ultimately, 182 questionnaires from 176 operators were evaluated. Because of incorrect or incomplete data, ca. 1/3 of the firms had to be contacted again by telephone or by e-mail. In most cases, we ultimately succeeded in obtaining consistent data sets. We did not ask about and consequently did not make any systematic balance sheet for chemicals added in treating the water (decarbonizing, flocculation, production of VE-water, regeneration of the ion-exchanger). Nonetheless, such consumption data were provided by some of the operators and were evaluated. Based on the annual consumption data for cooling water conditioners (on a productbasis) and the recipes for preparation in product information sheets, the loads of

30 defined ingredients and chemicals could be estimated. In cases where only an order of magnitude of the substance concentrations was documented, the following procedure was followed. Where less than or more than was given, the next lower or higher concentration after the decimal was used for the balance. Where concentration ranges were given, the mean was used. Example:

Given in the production information < 10% > 10% 10% - 20%

Assumed for the balance 9.9% 10.1% 15%

When no concentration was given, and only the product group was listed in the product information, typical concentrations from the literature or the available additive concentrations and product recipes were used. For example, for quarternary ammonium compounds and polycarboxylates no concentration could be obtained from the product information sheets. All data were subjected to a plausibility control. Thereby the annual water consumption per installed kW of cooling capacity was calculated as a basic parameter and the following classification was made: Cooling system Once-through cooling system Open recirculation cooling system Closed cooling system spec. water consumption [m 3/(kW*a)] 100 - 1000 1 - 100 0-1

Deviations from this rule of thumb classification indicated incorrect data or special features such as the use of hybrid cooling towers, a limited running time during the year or the like. In addition, the calculated concentrations of introduced substances in the wastewater of the individual plants were compared with the concentrations to be added as given in the literature (Anonymous 1994, Baltus and Berbee 1996, Fielden and Iddon 1997). For implausible concentrations the operators of the plants were contacted.

31

Estimating the annual loads for Germany: In order to calculate to an order of magnitude the total consumption of cooling water chemicals for Germany the questionnaires of the 176 plants were evaluated in separate categories of once-through, open recirculation, and closed cooling systems. The addition of chemicals in closed cooling systems was assumed to be irrelevant for the wastewater since the amounts added at the time of initial filling can only with some reservation be assigned to any years consumption (instead usually being disposed of as a concentrate in the garbage when the water is changed). Because of fundamental differences between industrial cooling systems and power plants (cf., sect. 5.4.4.1), the two categories were calculated separately. For the remaining cooling plants the percentage of plants that used a particular substance or a substance group was determined. Then the means and medians of the concentrations (on a substance basis rather than a product basis) were calculated from the annual water- and substance-consumption. Here the relationship of the amounts consumed to the added water volumes is a parameter which makes it possible to take into account the temporal components. (In principle, the substance consumption could also be derived from the wastewater volumes, but usually these were not known). For substances which were not continuously added, the calculated mean concentration generally lies well below the actual concentrations in practice. However, in individual cases, such as the shock treatment by batchwise addition of oxidative biocides to the circulation water, higher concentrations could also be calculated when the system volume is larger than the added water volume (cf., sect. 5.4.5). The estimation of the total loads of the non-continuously added chemicals (biocides) for Germany was then obtained from the following formula:

Fx = Concx * AX * Qx / 1000
where Fx Concx Ax Substance load [kg/a] Median of the balance concentration [mg/l] Relative proportion of the plants that use the substance

32 (e.g., 0.01 represents 1%) Qx x Water consumption for cooling purposes [m 3/a] Cooling principle (x = D for flow-through cooling and x = OK for open circulation cooling) This estimation is based on simplifying assumptions, which are briefly explained below: The classification of the plants as once-through, open recirculation, and closed cooling systems is not always unequivocal. There are substantial overlaps so as to give more of a continuum. The estimation of the total loads from the proportion of the plants that use this substance, based on the available data set provided by the operators, assumes that the specific consumption values are independent of the size of the cooling plant. This is only true to a certain extent. Thus, for example, larger plants tend to use oxidative and smaller plants non-oxidative biocides. However, because of the limited database at hand, a further sub-classification into various size classes, going beyond the separate consideration of industrial cooling and power plants, was not deemed appropriate. In general, it can be assumed that in the various industry branches, different requirements are placed on cooling water conditioning. Thus in the foodstuffs industry, because of hygiene requirements, a tendency to higher consumption amounts for biocides compared to other branches can be observed. However, the limited database does not permit a separate consideration of each branch. Although we asked the manufacturing firms to calculate the total loads on the basis of their product sales for cooling water conditioners in combination with the preparation recipes and the individual share of the market, we did not succeed in obtaining the desired results (Gartiser and Urich 2001). For continuously added conditioners (phosphonates, polycarboxylates) instead of the balance sheet values, the concentrations to be added in normal practice, as provided by industry, were used (cf., table 16). The annual loads of auxilliary additives (N-

33 methyl-2-pyrrolidone, alcohols) as well as inorganic pH-regulators and flocculation agents such as Fe(III)Cl3 for additional water treatment were not calculated for the Federal Republic of Germany, because the consumption in the plants was not systematically determined. 4.3 Literature and database-research The cooling water chemicals for which literature- and database-research was to be performed were determined on the basis of the product information. We did not consider chemicals for water treatment (inorganic acids and bases, salts for regeneration of the ion-exchanger) as well as auxilliary additives, e.g., solubilizing aids such as alcohols, which have no specific biocidal, dispersive or corrosion inhibiting effect. With the help of the following data banks/collections and databases, research was carried out on the individual substances: Substance data collections and fact databases Roth: Wassergefhrdende Stoffe, ecomed-Verlag, Landsberg Verschueren: Handbook of Environmental Data on Organic Chemicals, van Nostrand Reinhold (1996) Kommission zur Bewertung wassergefhrdender Stoffe (KBwS): Dokumentation wassergefhrdender Stoffe, Hirzel-Verlag, Stuttgart Merck Index 12th Ed. (1996) Ash: Handbook of Water Treatment Chemicals, Gower House (1996) Paulus: Microbicides for the Protection of Materials, Chapman & Hall (1993) Rossmoore: Handbook of Biocide and Preservative Use, Chapman & Hall (1995) Chemfinder, ECDIN and other Internet-Databases Dictionary of Substances and Their Effects (DOSE) International Uniform Chemical Information Database (IUCLID) Hazardous Substances Data Bank (HSDB) Data Bank of Environmental Properties of Chemicals (EnviChem) Register of Toxic Effects of Chemical Substances (RTECS)

34 Gefahrstoffinformationssystem (GESTIS) der gewerblichen Berufsgenossenschaften

PhysProp and Biolog/Biodeg (On-Line Databases of the Syracuse Research Cooperation, SRC, http://esc.syrres.com)

Literature databases Biological Abstracts Current Contents MEDLINE

The hazard risk statements (R-Phrases) in the dangerous substance regulations were researched in the GESTIS database. Here, in addition to the official classification according to Annex I of the Guideline 67/548/EWG, are also found selfevaluations by the producers (http://www.hvbg.de/d/bia/fac/zesp/zesp.htm). We went beyond this information and performed with varied success additional internetsearches of the databases of the U.S. Environmental Protection Agency (EPA), the U.S. National Institutes of Health (NIH) and the National Library of Medicine (including the database GENE-TOX). The available data were compared with the internal substance-database shared by Bund/Lnder (GSBL), which includes the UBA-Neustoffdatenbank for new substances, the databases of the KBwS (RIGOLETTO), the BgVV Chemis and BIG of the Feuerwehrinformationszentrum in Geel (Belgium). Additional information so obtained substantially extended our database. A preliminary listing of the results of our research was distributed to participating firms in preparation for the UBA-Workshop on the present project with the request that they fill in any gaps in the data (Gartiser and Urich 2001). As a result, the data pool was enlarged further. Whenever there were no citable published data available, the product-specific entries on the safety data sheets of the manufacturers were taken into account after consulting with them. The research results are presented in the Annex to this report. It remains to be noted that it was not the aim of this report to carry out a comprehensive and complete evaluation of all substances in the sense of a Risk Assessment in accordance with the laws on chemicals or biocide-

35 regulations. Consequently, as a standard for the selection of substance data we relied on the wastewater relevance of the data. In addition, we looked at the oral toxicity for mammals and the risk statements (R-Phrases), which are the prerequisite for classification in water-hazard classes (Anonymous 2000). Those organisms that are of importance in governmental control and are included in the list of parameters of the wastewater regulations were given a higher significance. Hereby it was sought to make a comparison possible with the practical investigations on wastewater sidestreams, products and active ingredients.

36

5 Results
5.1 Cooling water investigations The results of the investigated wastewater samples are presented in Table 6. A total of 12 cooling water samples and one condensate from steam production were tested. The pH of the cooling water samples lay between 7.9 and 9.5; the conductivity was between 121 s/cm (for VE-water in plant 2) and 10,130 S/cm (for the concentrate of the reverse-osmosis system in plant 1). The ecotoxicity of the cooling water samples in the algae, fluorescent bacteria and Daphnia tests was in most cases low (G A/L/D -values from 1 to 3). After a shocktreatment with quarternary ammonium compounds (plant 2) and isothiazolinones (plant 6), values up to a dilution factor of G L=196 were determined. However, after the elimination times of 1 to 3 days typically observed in practice, the ecotoxicity in the cooling water of plants 4, 7 and 2 was completely eliminated. In the cooling water of plant 5, 7 hours after the addition of QAV and hydrogen peroxide a slightly elevated ecotoxicity was measured, but this can be largely attributed to the input of solid materials during production. Also in the case of the continuous addition of bromochlorodimethylhydantoin, no ecotoxicity of the cooling water was observed. In the Ames test (screening) there was no mutagenic effect. However, one cooling water sample was toxic after a shock treatment with isothiazolinones (plant 6). The same sample turned out to be the only one that was genotoxic in the umu test (GEU=6).

37 Table 6: Summary of the wastewater investigations

Nr. 1 Electro industry KW KW Nr. 2 Plastic manufacturing Blow-down KW after Steam biocide shock after inacticondensate vation period dosage Nr. 3 Plastic manufacturing KW Nr. 4 Plastic film manufacturing KW KW Nr. 5 Plastic manufacturing KW KW 1 Nr. 6 Chemical industry Nr. 7 Chemical industry KW2 KW

Company No./ Branch

KW 2

Biocide (shock treatment)

(Isothia(QAV) zolinone)

QAV

(Isothiazolinones)

24 h after shock (Isothia- treatment H2O2/QAV zolinones) with Isothiazolino nes 12.07.00 8,8 720 14 0,8 1,0 1,0 0,9 1,5 1,5 2 2 1 17.07.01 8,6 831 11 1,1 0,8 1,1 1,0 1,5 1,5 1 2 1 27.07.00 8,7 240 302 (unfilt.) 138 (filtered) 0,9 1,0 1,1 0,7 1,5 1,5 32 (unfilt.) 4 384 (unfilt.) 2 (filtered)

BCDMH, (Bronopol, Isothiazolinones), NaOCl 11.12.00 8,3 735 <15 0,86 0,87 1,12 0,92 1,5 1,5 3 2 1

BCDMH, Bronopol, IsothiashockBCDMH zolinones dosage with Isothiazolinones 11.12.00 16.07.01 7,9 2170 138 tox. 1,04 tox. 1,07 6 6 64 192 24 7,8 2220 55 0,87 0,95 1,09 1,14 1,5 1,5 2 2 1 17.07.01 9,0 1620 23 1,01 0,93 1,09 1,1 1,5 1,5 1 2 1

Sampling pH Conductivity COD Ames test TA98 - S9 Ames test TA98 + S9 Ames test TA100 - S9 Ames test TA100 + S9 umu test - S9 umu test + S9 Algae toxicity Luminescent bacteria toxicity Daphnia toxicity (24 h) IF: Induction factor KW: Cooling water pH S/cm mg/l IF IF IF IF GEU GEU GA GL GD

27.06.00 19.07.00 9,5 10130 52 1,0 1,0 1,1 1,0 1,5 1,5 2 2 2 8,7 1640 24 1,0 1,3 1,1 0,9 3 1,5 2 2 1

11.10.00 8,60 2160 67 1,1 1,0 0,8 0,9 tox. tox. 192 48 96

14.10.00 9,00 121 50 1,0 0,7 1,1 1,1 1,5 1,5 2 2 1

19.07.00 7,6 31600 1721 (filtered) 0,9 1,0 1,1 1,1 1,5 1,5 4 2 4

19.07.00 8,1 130 17 1,0 1,2 1,1 1,0 1,5 1,5 2 2 1

38

5.2 Product investigations 5.2.1 Eco- and Genotoxicity The results on the mutagenicity, genotoxicity and ecotoxicity of the products used in the plants are presented in Table 7. Biocidal product 1, based on isothiazolinone, was found to be mutagenic or genotoxic in the Ames- and umu tests. The maximal induction factor was determined to be 11.1 (TA100). Biocidal product 2, based on a not further specified "heterocyclic sulfur/nitrogen compound," had a similar action spectrum and consequently is presumably also assignable to the isothiazolinone group. Biocidal product 3 also showed an increase in the induction factor in TA98 and in the umu test, although such an effect is not known for the group of the quarternary ammonium compounds (cf., annex); presumably, another component of this product is responsible. The effective concentrations (EC 50) in the fluorescent bacteria and Daphnia tests lie roughly in the range from 1 to 34 l/l; the corrosion inhibitor used in plant 1 turned out to be only slightly ecotoxic. 5.2.2 Identifying the source of the genotoxicity in plant 6 In the cooling water of circulation #2 of plant 6 a genotoxic effect was obtained in the umu test. At the time the sample was taken the cooling water was being continuously treated with 1-bromo-3-chloro-5,5-dimethylhydantoin (BCDMH) as well as with two additional products containing tolyltriazole, phosphonic acid and

nonylphenolethoxylate. On the day of sampling there had been an additional shock treatment with a combination preparation based on 2-bromo-2-nitropropan-1,3-diol (Bronopol) and isothiazolinone (5 liters to a 50 m 3 system volume). In order to assign the cause of the genotoxic effect, all these products were investigated in the umu test. Since no umu test results were available for Bronopol, it was also tested as the pure substance (Fluka-Chemika 17728). The results are shown in Table 8.

39 Table 7: Results of the product investigations

Luminescent bacteria test EC10 EC50 [l/l] [l/l] 1,15 3,74

Name

Usage in plant No.

Active substance

Dosage l/l

Ames test TA98 Maximum IF -S9 + S9 1,6

(20 l/Pl)

Ames test TA100 Maximum IF -S9 + S9 11,1

(20 l/Pl)

umu test Maximum IF -S9 + S9 3,2

(42 l/Pl)

Daphnia test EC10 [l/l] EC50 [l/l]

Biocide 1

Isothiazolinone

200-500

1,4
(20 l/Pl)

1,7
(4 l/Pl)

2,0
(42 l/Pl)

Corrosion inhibitor 1

NaOH, Triazoles

100-200 makeup water

1,2
(80 l/Pl)

1,7
(20 l/Pl)

1,5
(500 l/Pl)

1,3
(160 l/Pl

1,0

1,5
(333 l/Pl)

100-500 500-1000

Biocide 2

Heterocyclic sulfur/nitrogen compound QAV

100 shock dosage

1,3
(20 l/Pl)

1,9
(20 l/Pl)

10,1
(20 l/Pl)

1,5
(4 l/Pl)

2,9
(21 l/Pl)

2,9
(42 l/Pl)

0,6

2,3

1,6

Biocide 3

10-100

1,4

2,1

1,0

1,1

0,9
(5,2 l/Pl)

1,8
(42 l/Pl)

1,3

2,8

0,1

0,6

(100 l/Pl) (100 l/Pl)

Biocide 4

not known

100-250

1,5
(20 l/Pl)

1,7
(10 l/Pl)

1,4
(20 l/Pl)

1,0

1,6
(333 l/Pl)

1,3
(21 l/Pl)

5,8

29,5

0,2

34

Biocide 5

H2O2, QAV

100-200

1,6
(20 l/Pl)

1,8
(5 l/Pl)

1,2
(5 l/Pl)

1,0

1,6
(21 l/Pl)

1,6
(333 l/Pl)

0,6

2,5

0,8

1,8

IF: Induction factor

40 Table 8: Source of the genotoxicity in the cooling water from plant 6

Product InductionConcen- factor IF Active substance tration Growth-rate WR 1,5 1-Bromo-3-chloro50 mg/l 5,5-dimethylhydantoin Nonylphenolethoxylate Tolyltriazole Phosphonic acid 2-Bromo-2nitropropan-1,3diol (Bronopol) 2-Methyl-4isothiazolin-3-one 5-Chloro-2-methyl4-isothiazolin-3one Commercial chemical 2-Bromo-2nitropropan-1,3diol (Bronopol) 20 mg/l 50 l/l IF WR IF WR 200 l/l IF WR
Turbidity

Dilution factor G (- S9) 3 6 12 24 48 96 192 1,5

Dilution factor G (+ S9) 3 6 12 24 48 96 192

Biocide 6

0,53 0,54 0,78 0,77 1,00 1,00 1,11 0,96 0,69 0,87 0,93 0,85 0,84 1,05 0,96 0,90 1,38 1,35 1,23 1,24 0,95 1,04 0,91 1,12 1,13 1,01 1,01 1,04 1,05 0,97 0,92 0,93 0,84 0,77 1,03 0,94 0,93 0,82 0,85 1,00 0,76 0,83 0,82 0,98 0,78 1,17 1,24 0,89 1,04 1,06 1,00 1,01 1,00 1,19 1,13 1,07 0,98 1,05

Solubility aid for biocides Corrosion inhibitor 2

0,71 0,80 0,86 0,80 0,72 0,73 0,78 0,75 0,73 0,78 0,93 0,94 0,79 0,80 0,66 0,79 1,21 1,23 1,21 1,16 1,10 1,00 0,98 1,01 1,00 1,03 0,99 0,94 1,07 1,03 1,01 1,04

Biocide 7

100 l/l

IF WR

2,09 1,12 1,18 1,19 1,11 1,03 0,94 0,83 2,05 1,21 1,11 1,05 1,09 0,93 1,10 1,06

0,50 0,73 0,78 0,85 0,94 1,05 0,89 0,92 0,55 0,79 0,94 0,96 1,02 1,03 0,89 0,96

IF WR

tox.

tox.

2,63 1,78 1,35 1,04 1,24 1,02 tox.

2,20 1,53 1,26 1,13 1,05 1,09 1,06

0,15 0,39 0,58 0,87 0,99 0,98 0,88 0,93 0,39 0,73 0,88 0,97 1,01 1,01 0,99 0,94

41

It turned out that neither the biocide 1-bromo-3-chloro-5,5-dimethylhydantoin, the corrosion inhibitor, nor the solubilizing agent could explain the observed gentoxic effects. But biocide 7 proved to be weakly genotoxic, whereby the maximal induction factor of 2.0 was accompanied by a strong decrease in the growth factor (growth factor >0.5). From the product testing it is known that the isothiazolinones 2-methyl4-isothiazolin-3-one and 5-chloro-2-methyl-4-isothiazolin-3-one are genotoxic (cf., sect. 5.2.1). However, here induction factors up to 3.2 were determined. The substance testing of Bronopol yielded the result that this non-oxidative biocide also induced a weak effect in the umu test (up to an induction factor of 2.6), whereby also here the toxic effect must be taken into account in explaining the result. Thus, the genotoxic effect in circulation #2 of plant 6 is with high probability due to biocide 7, whereby the effect could be explained by both isothiazolinone as well as Bronopol. 5.2.3 Decrease of the biocidal effect in the fluorescent bacteria test 5.2.3.1 Starting concentration According to Annex 31 of the wastewater regulations, after a shock treatment with biocides the outflow may only be opened when for fluorescent bacteria toxicity the dilution step GL=12 is not exceeded. Since until now no generally accepted standard procedure was available for the determination of the elimination behavior of cooling water biocides in a batch experiment, various conditions with respect to the starting and inoculum concentration were simulated (cf., sect. 4.1.9). Our aim was to determine the influence of the biomass on the elimination factors degradation and adsorption. The starting concentration of the biocides was chosen based on the effective concentrations to be added according to values in the literature and manufacturers recommendations (cf., table 9).

42 Table 9: Dosing of biocides in the circulation cooling

Active substance

Concentration of active substance [mg/l]

3) 4) 0,1 - 0,2 1-5 0,2 - 5 0,02 - 0,2 1-5 1 - 50 1 - 50
200 - 1500 (?)

1) 2) Reference Oxidative biocides 0,1 - 0,2 Chlorine 5) Sodium hypochlorite 1-5 1-Bromo-3-chloro-5,52-7 0,2 - 5 dimethylhydantoin Ozone 0,015 - 0,2 0,015 - 0,2 Hydrogen peroxide 1-5 Non-oxidative biocides Glutardialdehyde 25 - 50 11 - 200 2,2-Dibromo-34 - 10 1 - 50 nitrilopropionamide 2-Bromo-2-nitropropan-1,31 - 25 1 - 25 diol (Bronopol) -Bromo--nitrostyrene 1-5 4 - 20 Isothiazolinone 1-5 1 - 30 Methylenebisthiocyanate 2-6 0,8 - 5 Quarternary ammonium 3 - 50 5 - 40 compounds
1) 2)

1-4 0,1 - 0,2 >50 25 - 50 4 - 10 1 - 25 1-5 1-2 2-6 3 - 15

1 - 30 0,75 - 5 6 - 10

Baltus and Berbee: Het gebruik van biociden in recirculatiekoelsysteme. RIZA, 1996 Consultants in Environmental Sciences: Biocides in Cooling Water Systems, April 1994 3) Fielden and Iddon: Use Category Document - Water Treatment Chemicals, Env.Agency, 1997 4) Personal communication Dr. Klautke, BetzDearborn GmbH, from 06.07.01 5) Required concentration without consideration of chlorine depletion

An overview of the experiments carried out and the most important results can be found in Table 10. In Figures 2-10 the elimination curves are presented.

43 Table 10: Experimental overview of the elimination curves of cooling water biocides
Concentration active substance mg/l Non-oxidative biocides 100 l/l (used Isothiazolinones concentration) 50 l/l Quarternary ammonium (used compounds concentration) Quarternary ammonium 100 l/l compounds with hydrogen (used peroxide concentration) Alkyldimethyl10 benzylammonium chloride 2,2-Dibromo-3nitrilopropionamide (DBNPA) Glutardialdehyde 10 48 30 60 120 160 35 Time until < 20% inhibition at dilution factor G12 0,5 g TS/l 30 mg TS/l cf., Figure No.

2d

>7d

3h

3h

not shown

3h 2d 48 h 7d

3h 3d 48 h >14 d 3d 3d 6d 7d >8 d

not shown 3 4 5 6

7d 4d

7 10

2-Bromo-2-nitro-propan-1,3diol (Bronopol) oxidative biocides 1-Bromo-3-chloro-5,5dimethylhydantoin TS: dry solids

4 37

1 h (G2) 44 h (G2)

1 h (G2) 44 h (G2)

8 9

5.2.3.2 Non-oxidative biocides Isothiazolinones The biocide based on isothiazolinones (biocide 2) was tested at the concentration to be used according to the manufacturers recommendation. There is a clearly discernible dependence on the density of the inoculum. While the fluorescent bacteria toxicity at the highest inoculum concentration was already completely reduced within one day, at the lowest concentration (post-settling outflow of a municipal sewage treatment plant) no elimination was observed even after 7 days (Figure 2). In a Daphnia test performed in parallel the product showed a similar elimination behavior at a dilution step of 12 (data not shown). Adsorption is a particularly important elimination factor here.

44

100 90

Isothiazolinone, Starting concentration 100 l product/l

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 20 40 60 80 100 120 140 160 180 1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung
TS: Dry solids; Nachklrung: Outflow final clarifier

Time [h]

Figure 2: Decrease of the fluorescent bacteria inhibition with isothiazolinone Quarternary ammonium compounds Two products with quarternary ammonium compounds (biocide 3 and biocide 4 as a combination product with hydrogen peroxide) were tested at the concentration recommended by the manufacturer. However, already at 3 h after the beginning of the test both products showed no further fluorescent bacteria toxicity (cf., table 10). Another product was tested at a defined concentration of 10 mg alkyl dimethylbenzyl ammonium chloride. Depending on the inoculum concentration, 2 typical curves were obtained. Above 200 mg d.s./l a 70% decrease in the fluorescent bacteria inhibition was already achieved 3 h after test begin, while in contrast, the time for a decrease at 30 mg d.s./l and in the outflow of the final-clarifier increased to 3 days (Figure 3). As a first approximation, biological de gradation and the adsorption of the quarternary ammonium chloride can be regarded as equally important elimination factors.

45

100 90

Alkyldimethylbenzylammonium chloride, Starting concentration 10 mg/l

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung

TS: Dry solids; Nachklrung: Outflow final clarifier

Time [h]

Figure 3: Decrease of the fluorescent bacteria inhibition with QAV

Dibromo-nitrilopropionamide (DBNPA) The biocide dibromo-nitrilopropionamide was investigated at two different

concentrations. At a concentration of 10 mg/l the fluorescent bacteria inhibition for all inoculum concentrations decreased rapidly within 3 h, but then more slowly thereafter. The threshhold of 20% inhibition was already reached after only 1 h in the experiment with the final-clarifier water. In contrast, in the tests with activated sludge up to 48 h were required (Figure 4). This difference is even more striking with the higher starting concentration of 48 mg/l; here the fluorescent bacteria inhibition in the final-clarifier water was completely eliminated within 24 h, while in tests with activated sludge at higher concentrations up to 14 days were required (Figure 5). The grounds for this can be found in the pH-dependence of the rate of hydrolysis of the compound. It is known that DBNPA is hydrolyzed substantially faster in the alkaline range than under neutral conditions. While the pH-value in the experiments with final-clarifier water lay between 7.6 and 8.2, in the tests with activated sludge a pH of 7.4 to 7.6 was measured (data not shown).

46

100 90

2,2-Dibromo-3-nitrilopropionamide (DBNPA), Starting concentration 10 mg/l

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 TS: Dry solids; Nachklrung: Outflow final clarifier 1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung

Time [h]

Figure 4: Decrease of the fluorescent bacteria inhibition with DBNPA (10 mg/l)

100 90

2,2-Dibromo-3-nitrilopropionamide (DBNPA), Starting concentration 48 mg/l

TS: Dry solids; Nachklrung: Outflow final clarifier

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 50 100 150 200

1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung

250

300

350

400

Time [h]

Figure 5: Decrease of the fluorescent bacteria inhibition with DBNPA (48 mg/l)

47 Glutaraldehyde Initially the influence of the glutaraldehyde concentration was investigated at a constant inoculum-content (30 mg d.s./l activated sludge). While the fluorescent bacteria inhibition at concentrations below 60 mg/l glutaraldehyde was fully eliminated within 3 days (Figure 6), the lag-phase at higher concentrations increased to 6 (at 120 mg/l) or 7 days (at 160 mg/l). In a second experiment the influence of the inoculum concentration at a constant concentration (160 mg/l) of glutraldehyde was determined. As was later noticed, this concentration was given in only 1 of 4 reference sources (Table 9). In the final-clarifier the fluorescent bacteria inhibition at this glutaraldeheyde concentration had not decreased after 7 days. At the highest inoculum concentration of 1 g d.s./l, in contrast, the inhibitory effect was completely eliminated after 3 days (Figure 7).

48

100 90

Glutardialdehyde; inoculum: activated sludge 30 mg dry solids/l

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 20 40 60 80 100 120 140 160 180 160 mg/l 120 mg/l 60 mg/l 30 mg/l

Time [h]

Figure 6: Fluorescent bacteria inhibition with glutardialdehyde (30-160 mg/l)

100 90

Glutardialdehyde, Starting concentration 160 mg/l

Inhibition at dilution factor 12 [%]

80 70 60 50 40 30 20 10 0 0 20 40 60 80 100 120 140 160 180 1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung

TS: Dry solids; Nachklrung: Outflow final clarifier

Time [h]

Figure 7: Fluorescent bacteria inhibition with glutardialdehyde (30-1000 mg d.s./l)

49 Bromonitropropandiol (Bronopol) The formaldehyde releasing biocide 2-bromo-2-nitro-propan-1,3-diol was

investigated at a concentration of 25 mg/l. The results showed a strict dependence of the elimination behavior on the inoculum density. At an inoculum concentration of 1 g d.s./l the fluorescent bacteria inhibition was largely eliminated wilthin 1 day, while at 30 mg d.s./l or in the outflow of the final-clarifier the inhibitory effect remained stable even after 8 days.

100 90

Inhibition at dilution factor 2 [%]

80 70 60 50 40 30 20 10 0 0 50 100

2-Bromo-2-nitro-propan-1,3-diol (Bronopol), Starting concentration 25 mg/l active substance

TS: Dry solids; Nachklrung: Outflow final clarifier

1000 mg TS/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung

150

200

250

Time [h]

Figure 8: Decrease of the fluorescent bacteria inhibition with Bronopol

50

5.2.3.3 Oxidative biocides Bromochlorodimethylhydantoin (BCDMH) Bromochlorodimethylhydantoin is used as a solid granulate with a sidestream water flow channeled through a filter filled with BCDMH. Because the granulate proved to be poorly soluble in the mineral salts nutrient solution needed for the Zahn-Wellens test, the substance was dissolved in tap water. The elimination behavior was determined for two starting concentrations. Because of the known instability of the compound, additional total chlorine determinations were carried out. At the required dilution step 12 there was no detectable fluorescent bacteria inhibition for either starting concentration, so the undiluted samples (G=2) were investigated. At a substance-concentration of 4 mg/l (1 mg total chlorine), for all inoculum concentrations there was no more measurable fluorescent bacteria toxicity after 3 h (Figure 8). At a substance-concentration of 37 mg/l (10 mg/l total chlorine) already within the first 6 h a clear dependence of elimination on the inoculum concentration could be observed. Thereafter, the inhibitory effect for all tests with inoculum lay between 10 and 30%, while the fluorescent bacteria inhibition in the test with tap water still lay at 90% even after 45 h (Figure 9). The decrease in the fluorescent bacteria toxicity accompanied the decrease in the total chlorine concentration (cf., table 11). Table 11: Elimination of BCDMH depending on the inoculum

Free chlorine [mg/l]

0h 3h 20 h 44 h 500 mg TS/l 2,08 1,11 0,12 0,11 200 mg TS/l 0,48 0,13 0,05 0,07 30 mg Outflow final TS/l clarifier 0,17 0,98 0,15 0,49 0,10 0,30 0,05 0,04 Tap water 4,98 5,55 4,78 0,57 Tap water 8,30 8,65 6,45 0,61

Total chlorine [mg/l]

500 mg 200 mg 30 mg Outflow final TS/l TS/l TS/l clarifier 0h 4,92 3,81 2,62 4,09 3h 2,70 2,48 0,82 2,92 20 h 0,72 0,49 0,17 1,13 44 h 0,14 0,09 0,07 0,07 Start concentration 9,9 mg/l total chlorine

51

100 90

1-Bromo-3-chloro-5,5-dimethylhydantoin, Starting concentration 1 mg/l total chlorine

Inhibition at dilution factor 2 [%]

80
TS: Dry solids; Nachklrung: Outflow final clarifier; Leitungswasser: Tap water

70 60 50 40 30 20 10 0 0 0,5 1 1,5 2 2,5 3 3,5 1000 mg/l 500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung Leitungswasser

Time [h]

Figure 9: Decrease of the fluorescent bacteria inhibition with BCDMH (4 mg/l)

120 110 100

1-Bromo-3-chloro-5,5-dimethylhydantoin, Starting concentration 9.9 mg/l total chlorine

Inhibition at dilution factor 2 [%]

90 80 70 60 50 40 30 20 10 0 0 5 10 15 20 25 30 35 40 45 50
TS: Dry solids; Nachklrung: Outflow final clarifier; Leitungswasser: Tap water

500 mg TS/l 200 mg TS/l 30 mg TS/l Nachklrung Leitungswasser

Time [h]

Figure 10: Decrease of the fluorescent bacteria inhibition with BCDMH (37 mg/l)

52 5.3 Evaluation of the product information sheets In the course of the study documentation was compiled for a total of 418 products for cooling water conditioning from ca. 35 manufacturing firms. A total of 101 products therefrom contained biocides. The qualitity of the documents ranged from poorly documented safety data sheets to detailed eco logical evaluations. Very few elimination curves for cooling water biocides were provided. Since all the leading producers in the field sent us information, an estimated 80% of the total market has been surveyed. In Figure 11 the product groups for the cooling water conditioners are documented.

Not known 3% Others 3% Guanidines 3% Quaternary ammonium compounds 9% Organic sulfuric compounds 8% Oxygen releasers 8% Aldehydes 6% Isothiazolinones 26% Organic bromine compounds 10% Chlorine releasers 8% Bromine releasers 16%

Figure 11: Proportion of biocidal active ingredients in 101 products

53

5.4 Evaluation of the questionnnaires Altogether 182 questionnaires on the cooling systems of 176 plants were evaluated. Of these, 110 had open circulation cooling, 48 had once-through cooling, and 24 could be classified as closed-circulation cooling. The most important characteristic data from the data set are given in Table 12. Table 12: Characteristic data for the cooling systems investigated
Cooling systems with Cooling Cooling Water systems capacity consumption [n] Open recirculation Industrial cooling cooling systems Once-through cooling 48 Closed circulation cooling Sum
[n]: Number

Mean specific water consumption [m /(kW*a)] 6

3

Products used [t/a] 83

Biocides

Condition- Biocides and ing conditioning [n] [n] 8 4

MW 10 4.817

[1000 m /a] 40.380

[n] 5

Power plants

100

2.051

35.421

20

778

69

72

62

18.936 198 26.002

5.985.084 13 6.060.898

501 0,6

23 7
*)

4 4 82

4 16 100

2 4 72

24 182

891

*) Partly once-only filling not attributable to annual amount

The water consumption for cooling purposes of ca. 6 billion m 3 represents 19% of the total untreated discharged cooling water (Stat. Bundesamt 1998). Only 76 Mio. m 3 are assigned to open circulation cooling. Overall, it could be established that in most cases it was not possible to make an unequivocal assignment of the continuously added substances as well as their concentration ranges from the product information; often only the substance group was indicated. For the batchwise added biocides, in contrast, the names of the active substances and their concentration ranges are better documented, because these are listed as hazardous substances in the safety information sheets.

54 5.4.1 Open recirculation cooling systems In the cooling systems investigated a total of 891 t of cooling water conditioners (weight on a product basis) was used. This does not include the chemicals used for treatment of raw water (de-carbonizing, flocculation and pH-adjustment) such as inorganic acids and bases. For these latter purposes, a total of 53 t of sulfuric acid, 4 t of hydrochloric acid, 0.2 t of sodium hydroxide, 145 t of ferric(III) chloride (weights of the pure substances) as well as 2 t of organic flocculation aids (weight of the products) were consumed. The total loads of defined substances which were used in the 110 plants for open circulation cooling are given in Table 13. Here products were not considered if they could not be assigned to a group or if the concentrations used were not given, so the values given are minimal values. For the 10 power plants with open circulation cooling investigated, no zinc or molybdate was used, in agreement with Annex 31. Therefore, it seems justified to consider these separately from the industrial cooling systems.

55 Table 13: Total amounts of the investigated chemicals in open circulation cooling
Industrial cooling systems [t/a] 1) Oxidative biocides Chlorine Sodium hypochlorite Calcium hypochlorite Trichlorisocyanuric acid Ammonium bromide / sodium bromide Sodium hypobromite 1-Bromo-3-chloro-5,5dimethylhydantoin Hydrogen peroxide Ozone Potassium peroxymonobisulfate Peracetic acid Non oxidative biocides 2,2-Dibromo-3-nitrilopropionamide 2-Bromo-2-nitropropan-1,3-diol (Bronopol) Isothiazolinones S-Triazine Dodecylguanidine hydrochloride (DGH) Methylenebisthiocyanate 52,3 27,7 0,01 0,03 ? 0,5 1,6 9,7 0,2 0,1 0,3 0,7

Power plants [t/a] 2)

8,2

0,1 0,01 0,6 ? 0,01 0,1

Quaternary ammonium compounds ? Conditioning chemicals Phosphonic acids 6,6 1,7 Phosphates 0,7 Triazoles 2,4 ? Polycarboxylates >7,7 >0,3 Sodium molybdate 1,0 Zinc salts 11,9 Auxiliary additives/water treatment N-methyl-2-pyrrolidone 0,2 Alcohols 2,4 Fe(III)Cl3 145,2 inorganic lyes 1,6 0,1 inorganic acids 42,2 4,9 1) 100 industrial cooling systems with open recirculation cooling 2) 10 power plants with open recirculation cooling

56 5.4.2 Once-through cooling systems The 48 investigated plants with once-through cooling systems include about 20% each from the branches foodstuffs industry, chemical industry and energy production, with the remaining plants belonging to the plastics, metal and paper industries, as well as other areas. Only 6 of the 48 plants used cooling water conditioners. These plants should be considered separately (see Table 14). In one food processing plant the canned goods were cooled with water after sterilization. In order to prevent contamination or corrosion of the cans, fatty alcohol ethoxylates, sodium ethyl-hexanoate and chlorine in the form of sodium hypochlorite were continuously added to the cooling water. In another plant manufacturing canned foodstuffs, corrosion inhibitors (sodium molybdate; alcohol ethoxylates) and hardness stabilizers (phosphocarbonic acids) were continuously fed into the cooling water. In one plant, which manufactures packaging materials, the cooling water is treated with isothiazolinone and phosphonic acids. The provision of cooling waters here is made more difficult by the very high hardness (25 dH) and manganese content. In one plant of the chemical industry chlorine was continuously added to the cooling water. Sodium hypochlorite is used in a thermal power plant cooled by sea water. In one large thermal power plant iron(II) sulfate is used, since in low concentrations (0.05-1 mg/l) in sea water and brackish water it acts anticorrosively through the formation of a protective layer of iron hydroxide (Wunderlich 1978a). All in all, the use and selection of conditioners in plants with once-though water cooling presents a very heterogeneous picture.

57

Table 14: Consumption data in plants with flow-through cooling

Nr. Branch Capacity [MW] Water consumption [1000 m /a] [m /(a*KW)]
3 3

Origin of water

Biocide Product dosage consumption [kg]

Discharge

Active ingredient 1 [kg] Sodium molybdate dihydrate

Active ingredient 2 [kg]

Active ingredient 3 [kg]

Active ingredient 4 [kg]

37 Foodstuffs industry

17

other Product 1 none

2.400

KA

Cocoimino132,0 propionate Sodium 2ethylhexano 258,7 ate 50,7 NaOH 300,0 1560,0

Alkanol132,0 ethoxylate

Phosphoncarbonic 132 acids

132

65 Foodstuffs industry 68 Power plant Power 83 plant

314 240 420.000 124

O G M

3.400

1.486

1.040.000

700

(M)

Product 2 Product 3 Chlorine Sodium hypochlorite Ferrous(II)sulfate

S K S

O and Ethoxylated 1.300 KA alcohols NaOCl 502 Chlorine 300 NaOCl 12.000 M

63,7 49,7

450.000

Ferrous(II)sulfate

120 Bandage material

0,2

55

275

Product 4

430

5-Chlor-2-methyl4-isothiazolin-3one

Product 5 134 Chemical ? 7.000 O Chlorine K industry Origin of water: G=ground water; O=surface water; M= marine water; (M)=brackish water Biocide dosage: K=continuously, S=batchwise (shock) dosing Discharge: O=surface water; KA=municipal treatment plant; M=sea

1.200 4.380

Phosphonic acids Chlorine

2-Methyl-4isothiazolin6,5 3-one Butanephosphonotricarbonic acid 4380,0

2,2

58

5.4.3 Closed circulation cooling systems The 24 closed circulation cooling systems investigated were mostly small plalnts with less than 10 MW cooling capacity. In 8 plants the cooling water is not conditioned. In all, the chemicals added weighed 13 t (on a product basis), but only a small proportion (< 0.5 t) could be attributed to defined chemicals. A large proportion of the products consists of not further characterized antifreeze agents. To inhibit corrosion, molybdate and zinc salts are used, as well as phosphonic acids and triazoles. As biocides, quarternary ammonium compounds, organobromine compounds and isothiazolinone are used. The estimation of the total amounts used was made even more difficult by the fact that in some cases the first filling of the closed systems was involved and as a rule the circulation water is only changed at intervals of several years. According to the operators responses on the filled out questionnaires, ca. 1/3 of the plants have the cooling water taken away for disposal after use, while 2/3 of the plants release it into the municipal sewage treatment plants, in individual cases also over the surface water. 5.4.4 Estimation of the total loads for Germany 5.4.4.1 Water consumption for cooling purposes In keeping with our originally stated aims, on the basis of the consumption data of the cooling plants surveyed an overall balance sheet calculation is to be made of the total annual loads for conditioners (on a substance basis) for Germany as a whole. Here the water consumption data for cooling purposes collected by the German Federal Statistics Office is directly applied (Stat. Bundesamt 1998). However, the Statistics Office does not distinguish between plants with flow-through cooling and those with open circulation cooling, so that several assumptions must be made. a) Mining and processing activities In 1995 376 Mio. m 3 of cooling water were discharged after recirculation cooling and 4.981 Mio. m3 without recirculation cooling (Stat. Bundesamt 1998). While in a first approximation the discharged amount without recirculation cooling equals the amount of water used, for the plants with recirculation cooling the evaporation losses must be considered. For a concentration factor of 3 ca. 1/3 of the make-up water is re-emitted, while 2/3 evaporates. Thus a total water use for recirculation cooling

59 plants of 1.1 billion m 3 (109 m3) can be calculated. However, hereby once-through cooling systems with outflow cooling are included. An alternative procedure is based on the data given by the Statistics Office for the total use of cooling waters and the utilization factor for multiple and recirculation use (Stat. Bundesamt 1998). In calculating the total use the recirculation water is included repeatedly with a factor corresponding to the number of times it recirculates. "Multiple use" is understood to mean the further use of the cooling waters for other purposes (e.g., for cleaning). From the total use of the cooling water (29.1 billion m 3) and the utilization factor of 4.8 (Statistisches Bundesamt, Fachserie 19, Reihe 2.2 (1995), 3.1) it can be calculated that ca. 6 billion m 3 fresh water were used for cooling purposes in 1995. If one subtracts from this value the amount of water used only once (i.e., without recirculation cooling), then ca. 1 billion m 3 water were used for open recirculation cooling and 5 billion m 3 water for once-through circulation cooling. By and large, these data agree with the rule of thumb numerical classification given above. Using these values we then performed the calculations to obtain the overall estimates. b) Cooling water provision for the thermal power plants for the public power supply The thermal power plants supplying the public have a special position and are treated separately by the Federal Statistics Office and also in Annex 31. The operations of the plants are more uniform in comparison to the industrial cooling systems. The principal demand for cooling water is for the turbine condensors, and secondary cooling locations are usually operated in closed systems. For large cooling systems of modern power plants with condensors made of stainless steel or titanium the addition of corrosion inhibitors is not necessary. Large cooling towers of power plants should be so constructed and operated that the regular use of microbiocides is not necessary (Fichte et al. 2000). The total amount of cooling water used in 1995 (including multiple use and circulation use) was ca. 61 billion m 3, and the utilization factor was ca. 2.2 (Stat. Bundesamt 1998). In response to our enquiry, the Federal Statistics Office provided us with the data for the proportion of multiple use and circulation use, as this could not be ascertained from the published information (Mr. Knichel 2001). From the total amount of discharged cooling water, by subtracting the water used only once, the proportion used for multiple use and

60 circulation cooling water was calculated in the first approximation to be 4% of the total discharged cooling water (cf., table 15). Table 15: Cooling water use for thermal power plants (1995 )

Multiple use Circulation use Single use Sums

Cooling water used [1000 m3] 458.861 34.354.663 26.946.470 61.759.994

Cooling water discharged [%] [1000 m3] 1.126.255 26.946.470 28.072.725 4 96 100

Sources: (Mr. Knichel 2001); (Stat. Bundesamt 1998) Based on these considerations, it was assumed for further calculations that 1 billion m3 cooling water are discharged from open circulation cooling and 27 billion m 3 cooling water from once-through cooling. Assuming a concentration factor of 2, this represents ca. 2 billion m 3/a of fresh water consumed for open circulation cooling of power plants. 5.4.4.2 Open circulation cooling The procedure for calculating the concentrations of chemicals and the proportion (percentage) of the plants which use a particular chemical or chemical group is described in Sect. 4.2.3. The total loads are obtained from the total amount of 1 billion m3 cooling water used for mining and processing industries. The thermal power plants, which use ca. 2 billion m 3 of additional water, were calculated separately. Continuously added conditioners A preliminary version of the balance drawn up from these data was sent to the participants of the workshop of 18.05.01 for their comments. Several participants expressed skepticism about the level in the balance for the added concentrations and loads, especially for the continuously added cooling water conditioners such as the phosphonates. In response, an accounting for this compound was additionally undertaken on the basis of the concentrations added in normal practice (cf., table 16) Since the standard concentrations are based on the circulation water, the estimation of the total load here is not made on the basis of the fresh water volumes, but rather

61 on the actually discharged cooling water (376 Mio. m 3 from industrial cooling plants, 1 billion m3 from power plants). The consumption of auxiliary aids (N-methyl-2pyrrolidone, alcohols) as well as inorganic pH-regulators and flocculants like Fe(III)Cl3 was not systematically surveyed and thus not included in the calculation for total consumption for all of Germany. In the substance group polycarboxylates several differently named compounds were included (polycarboxylate, polycarbonic acid, alkylepoxycarboxylate, polyoxycarbonic acid, polymaleic acid, maleic acid/copolymer, acrylic acid-co-polymer, acrylic acid-Ter-polymer). The less important low molecular weight acids were omitted here. Table 16: Concentrations of continuously added conditioners (concentration in the circulation water in mg/l)
Active substance Phosphonate (~ ppm P) 6) Phosphate (~ ppm P) Polycarboxylates and Polycarbonic Zinc Molybdate Triazole derivatives
1) 2)

Industrial coolingPower plants 2), 3) water circulation 3-8 (~ 0.8 - 2) 12 - 15 (~ 4 - 5) 6 - 12 1-3 5-7 1-2 1.5 - 3 (~ 0.4 - 0.8) 6 (~ 2) 2-7 1-2

Corrosion inhibitors 4) 5 - 20 (~ 1.3 - 5) 5 - 25 (~ 1.6 - 8) 2 - 10 2 - 15 1-5

Hardness Cooling water 5 stabilizers/Dispersants 4) ) 2 - 10 (~ 0.5 - 0.8) n.a. 5 - 20 6 - 12 (~ 1.5 - 3) 7 - 15 (~ 2.2 - 4.8) 20 - 50 3-4
n.a. n. a.

n.a.= no information available; "-" = not used for this purpose Personal communication from Dr. Klautke, BetzDearborn GmbH , 09.01.02 Personal communication from Dr. Olkis, ONDEO Nalco GmbH, 10.12.01 3) Personal communication from Dr. Hater, Henkel Surface Technologies, 13.12.02 4) Fielden et al. (1997) 5) Wunderlich (1978) 6) The usual P-content of phosphonates lies between 10 and 30%, due to the substances present in the largest amounts, HEDP and ATMP, with 30% P-content assumed to be present at an average content of 25%.

Shock treatments by batchwise addition of cooling water biocides The balance sheet calculations used for continuously added chemicals could not be applied for the biocides which are usually added batchwise, as this would have led to a gross overestimation of the loads. In order to take into account the time factor, the total loads were calculated as described in sect. 4.2.3 from the consumption numbers and the documented fresh water use.The concentration calculated here represents a helpful parameter, which does not necessarily correspond to the concentration in the circulation water (after a shock treatment).

62 For the biocide groups of quarternary ammonium compounds (QAV), triazine derivatives and guanidine, no information on concentrations was available for the products involved. It was assumed that these biocides as well as isothiazolinones were usually added batchwise at long intervals. The median of the calculated substance concentration for isothiazolinones is 0.08 mg/l (cf., table 17). To take into account the higher concentrations used for QAV and triazine derivatives this value was adjusted upward, to 0.26 mg/l (QAV) and 0.533 mg/l (triazine/guanidine). Here it was assumed that the effective substance concentrations of isothiazolinones, QAV and triazine/guanidine correspond to 3, 10 and 20 mg/l (cf., also table 9). In Table 17 the balance of the total amounts of cooling water chemicals in open circulation systems are presented for Germany, with industrial cooling systems and power plants given separately. For continuously added conditioners the loads are calculated in two ways for comparison, from the concentrations usually applied in practice and from the concentrations inferred from drawing up the balance.

63

Table 17: Chemical usage in open circulation cooling systems in Germany

Industrial cooling systems
Median of active Proportion subst. of plants concentration [%] [mg/l] Oxidative biocides 1) Chlorine Sodium hypochlorite Calcium hypochlorite Trichlorisocyanuric acid Ammonium bromide / sodium bromide Sodium hypobromite 1-Brom-3-chloro-5,5dimethylhydantoin Hydrogen peroxide Ozone Potassium peroxymonobisulfate Peracetic acid 1) Non-oxidative biocides 2,2-Dibromo-3-nitrilopropionamide 2-Bromo-2-nitropropan-1,3-diol (Bronopol) Isothiazolinones S-Triazines Dodecylguanidine hydrochloride (DGH) Methylenebisthiocyanate 4,6 2,4 14,6 1 ? 1,1 3,4 23,6 0,3 1,1 5 4 11 1 1 2 4 Load [t/a]

Power plants
Median of Proportion active subst. of plants concentration [%] [mg/l] Load [t/a]

Sum Total load [t/a]

184 264 (146) (10)

2,05

10

(410)

184 (674) (146) (10) ? (44)

(44) 3,64 20 (1456)

11 374 5 (1180) 1 (3) 1 (11) 1 (50)

(1830) (1180) (3) (11) (50)

0,66 0,02 0,08 0,53 0,69 0,08

3 2 26 2 1 3

20 0,4 21 11 (7) 2 11 421 948 124 1218 135 113 160 1932 242 ? 45 216
3

20 0,4 21 (111) (7) 2 0,267 6 2,3 1,5 4,5 0,4 ? ? 10 0 30 30 10 0 30 30 10 (53) 0 690 450 (450) (64) 421 1638 574 (1668) 135 113 160 2172 242 45 216

0,5

10

(100)

Quaternary ammonium compounds 0,267 4 Conditioning chemicals, concentrations used in practice 1) Phosphate 14 8 Phosphonic acids 6 42 Triazoles 1,5 22 Polycarboxylates 9 36 Molybdate 6 6 Zinc 2 15 Conditioning chemicals, balanced concentration 1) Phosphate 2 8 Phosphonic acids 4,6 42 Triazoles 1,1 22 Polycarboxylates ? 36 Molybdate 0,752 6 Zinc 1,44 15
3

240

1) With reference to the fresh water consumption (1000 Mill. m for industrial ccoling, 2000 Mill. m for power plants) 3 3 2) With reference to the discharged water (376 Mill. m for industrial systems, 1000 Mill. m for power plants) Values in brackets are based on unreliable data basis

5.4.4.3 Once-through cooling The database for an estimation of the total loads of chemicals used in once-through cooling systems is not adequate for a reliable calculation, because in most plants conditioners are not used. An extrapolation of total loads from the 6 plants given (cf., table 14) is very unreliable, since these represent individual examples.

64 At best, an evaluation of the foodstuffs industry sector seems permissible, although even these numbers have only limited validity (Table 18). In the foodstuffs industry, based on the data documented in Table 3, ca. 3.2% of the cooling water is consumed in the processing industry. This corresponds to ca 160 Mio. m 3 cooling water for once-through cooling (3.2% of 5 billion m 3). In one thermal power plant 1.6 t/a active chlorine is added for the treatment of sea water from the Baltic. However, this value is not representative for once-through water cooled thermal power plants. According to an estimate made by the Bund/Lnder working group on cooling water ca. 50 t/a of chlorine are used in thermal power plants with once-through cooling (Piegsa 2001a). Only in one plant of the chemical industry is chlorine continuously added. In the discharge of this plant, according ot the operator, there is no detectable chlorine and the AOX-value of ca. 50 g/l lies substantially below the threshhold value. Table 18: Total load of cooling water chemicals in the once-through cooling of the foodstuffs industry
Chemical Proportion of concentration companies [mg/l] [%] 7,8 8 0,6 8 0,5 8 7,8 8 7,8 8
3

Total load 1) [t/a] 104 8 6 104 104

Sodium molybdate dihydrate Chlorine and chlorine releasers Ethoxylated alcohols Cocoiminopropionate Phosphono-carbonic acids
1)

Cooling water demand of the branch 160 Mill. m

5.4.5 Overview and comparison The balance sheet of the chemical usage in open circulation cooling shows a consumption of ca. 4100 t/a for oxidative and 125 t/a for non-oxidative biocides. For once-through cooling systems there are, in addition, ca. 50 t of chlorine/chlorinereleasers. The balance shows chlorine concentrations in 4 plants with open circulation cooling of ca. 4 to 5 mg/l, substantially higher than the value of 0.2 mg/l given in Table 9. However, the concentration added depends primarily on the chlorine removal, and the value of 0.3 mg/l pertains to the targeted residual chlorine concentration in the circulation water. The balance for the chlorine concentration of 4.6 mg/l was based on data reports from only 3 operators, so these values are quite

65 possibly non-representative. The sodium hypochlorite-concentration (2.1-2.4 mg/l active chlorine) was, in contrast, calculated from 10 reports and is thus more valid. However, more extensive research on the cooling plants in which chlorine was used revealed that the hydraulic conditions also had to be considered. As examples, 2 cases are described:
In one plant of the chemical industry with 2 open cooling systems (Cycles of concentration 2.5 3.0) 100 kg of liquid chlorine are added batchwise daily to a total circulating volume of 48.000 m /h and the outflow is closed for 4 h. The measured chlorine concentration after the shock treatment is 0.3 to 0.8 mg/l. However, if one relates the annual consumption of active chlorine of 30.000 kg/a to the amount of fresh water used of 6.5 Mio. m /a, the hypothetical concentration "in the inflow" is increased to 4.6 mg/l. The ground for this lies in the fact that the amount of fresh make-up water introduced is less than the system volume (which is not known to the operator). In another plant, also from the chemical industry, ca. 20 kg of liquid chlorine is automatically added daily to each of 3 cooling circulation systems (concentration factor 2.0 - 2.5) , until a free chlorine content in the circulation water of 0.3 mg/l is reached (corresponding to a total amount added of 2.9 mg/l of circulation water). The overall system volume is 21.000 m , and the mean amount of make-up water added is 9.378 m /d. From this it can be calculated that the system volume (neglecting the evaporation) is only exchanged once every 2 days. Thereby the hypothetical chlorine concentration "in the inflow" increases to 6 mg/l.
3 3 3 3

Thus for batchwise addition of chemicals to the circulation water when the consumption is based on the volume of added make-up water, substantially higher hypothetical concentrations may be calculated than the corresponding applied concentration. While the consumption of sodium hypochlorite (including chlorine) in Germany is at about the same level as in the neighboring European countries, the consumption of hydrogen peroxide and peroxyacetic acid in Germany is notably higher (cf., table 19, after Anonymous 2000). However, the loads for Germany have been extrapolated from only 5 reports on the use of H 2O2 and one mention of the use of peroxyacetic acid. Hydrogen peroxide is used at high substance concentrations (up to 50 mg/l) and is frequently added continuously, so the large loads seem plausible at first inspection. However, the calculated consumption numbers for Germany are probably too high, since hydrogen peroxide is for the most part used in small to medium-sized cooling systems. A restriction of its use to selected branches of industry, as supposed by certain employees of the manufacturers, could not be confirmed.

66 It was noted by the producers that the loads calculated for BCDMH are also too high, because this biocide is primarily used in smaller plants. The balance drawn up revealed that BCDM was used in 13 of 110 plants with open circulation cooling with a total cooling capacity of 140 MW and a total water consumption of 1.5 Mio m 3. In the largest plant (chemical industry), comprising 2 cooling towers, for a total capacity of 50 MW and a water consumption of 840,000 m 3, ca. 750 kg/a of BCDMH were consumed. Both power plants were small thermal power plants, with a cooling capacity under 25 MW. The production of sodium hypobromite generally takes place on site, through the reaction of sodium bromide with sodium hypochlorite. The consumption of sodium hypochlorite was taken into account in Table 17, but for sodium bromide no concentration data were available. The total load of sodium hypobromite in the balance in Table 17 is based on one product based with stabilized sodium hypobromite. The extrapolation of the consumption data for calcium hypochlorite is just as inaccurate. Its use (as a solid) in cooling systems is unusual, but is documented in one actual case. For non-oxidative biocides the consumption data for quarternary ammonium compounds, isothiazolinones, dibromonitrilopropionamide and methylene-

(bis)thiocyanate are comparable with the values for England. For some biocides such as -bromo--nitrostyrene or glutardialdehyde no consumption values for Germany were provided, although products with these active ingredients are on the market. On the other hand, other biocides such as triazine derivatives, guanidine and Bronopol, for which relevant loads were determined in this study, were not investigated in neighboring countries. The total load for triazines was substantially influenced by its rather untypical use in one of the surveyed power plants, and the high water consumption of power plants over all, which appears unrealistically high. In a first approximation, the total consumption should be maximally of the same order of magnitude as for the other non-oxidative biocides (ca. 10-20 t/a). For hardness stabilisation and corrosion inhibition ca. 1,600-2,200 t of phosphonic acids were consumed in open circulation cooling. In an earlier study a consumption of ca. 3.900 t of phosphonates was reported for Europe (without indicating the year) for cooling water conditioning (Gledhill and Feijtel 1992, Meerkerk and Puijker 1997). Horstmann et al. (1986) estimated the phosphonate load from cooling circulation for

67 Germany (the old Bundeslnder) to be 3,400 t/a, assuming for all cooling water circulation a phosphonate concentration of 5 mg/l and for indirect dischargers a 50% elimination of the phosphonates (Horstmann and Grohmann 1986). Since in this estimate the concentration factor was not taken into account, the number was later corrected to 1,200 t/a (Hsselbarth 1987). The amount of PBTC produced in Germany is given as 1,000-10,000 t/a, and consumption is estimated to be 5001,000 t/a. Of this amount, 40% is used in cooling systems (UNEP 1996). Thus the total load of phosphonates calculated in this project lies at a realistic level. (The total consumption of phosphonates in all areas of application for the entire U.S.A. is estimated to be ca. 20,000 t/a; cf., Nowack and Stone 1999). Knepper et al. (2002) used the data in the Register for washing detergents and cleaning agents to estimate the consumption of ATMP, HEDP, EDTMP, PBTC and DTPMP as complex-formers for this purpose in Germany at 11.100 t/a. For the areas of textile-, leather- and paper-aids another 472 t/a are used. However, these values are also inaccurate because they fail to include changes in the formulations to be found in the database and are probably substantial overestimates (Kleinstck 2002). Heavy metals used in open circulation cooling include ca. 45-135 t molybdenum in the form of sodium molybdate and 110-220 t zinc, mostly in the form of zinc chloride. On the other hand, a survey of the emissions of zinc from the cooling water area for Nordrhein-Westfalen gave a total load of only 5 t zinc/a from cooling water emissions for the drainage basins of the Rhine, Lippe and Ruhr Rivers (Piegsa 2001a). However, there are altogether only very few comparative numbers which could be used to evaluate the reliability of the balance drawn up. Wunderlich (1978) estimated the consumption of hardness stabilizers, corrosion inhibitors and dispersents in cooling water conditioning at 12,000 to 20,000 t/a (on a substance-weight basis). To this total, the biocides must be added, whereby 95% of the biocides were used in circulation cooling. The applied biocides included chlorine and inorganic chlorine compounds (95%), acrolein (1%), quarternary ammonium compounds (2%) and others (together 2%, Wunderlich 1978a). The estimate was based on the then available cooling systems (flow-through and open circulation cooling), as well as the concentration of substances corresponding to the most likely used doses and procedures (Wunderlich 2002).

68 In principle the size of the cooling plants could also be considered in estimating total loads of cooling water conditioners, so as not to overestimate the load of, for example, biocides like hydrogen peroxide or BCDMH, which are primarily used in smaller plants. However, data on the size in terms of the water volumes are only available for the 314 thermal power plants supplying the public (from < 5 Mio m 3/a to >500 Mio. m3/a; cf., Stat. Bundesamt 1998). So as to preserve the systematic basis of and the plausibility or validity of the balance, it was consciously decided not to attempt to include any, at best rather arbitrary, size factor in assuming the total water consumption. In order to introduce such a factor, the da tabase would have to be substantially enlarged. Instead, doubtful data in the balance data in Table 17 are suitably marked.

69 Table 19: Comparison of the estimated consumption data for certain biocides with data from other countries (data in kg/a on a substance basis)

Active substance
Chlorine based Chlorine Sodium hypochlorite Calcium hypochlorite Sodium dichlorisocyanuric acid Chlorodioxide Bromine based Sodium bromide 1-Brom-3-chlor-5,5dimethylhydantoin (BCDMH) Sodium hypobromite Ozone Hydrogen peroxide Potassium peroxymonobisulfate Peracetic acid

UK

1)

NL

1)

F
817.000

2)

oxidative biocides
731.000 19.300 13.000 356.000 286.000 0 910 2.100.000 184.000 674.000 146.000 10.000

1.830.000 44.000 3.000 1.180.000 11.000 50.000

975

Total amount of oxidative biocides 1.407.185 2.100.000 817.000 4.132.000 non-oxidative biocides
Dimethylcocobenzyl-ammonium chloride Benzylalkylammonium chloride Total amount of QAV 5-Chlor-2-methyl-4-isothiazolin-3Isothiazolinones on QAV Total amount of Isothiazolinones halogenated Bisphenols (Dichlorophen, Fentichlor) Thiocarbamate Glutardialdehyde Tetraalkylphosphoniumchloride 2,2-Dibromo-3-nitrilopropionamid 2-Bromo-2-nitropropan-1,3-diol S-Triazine Dodecylguanidinehydrochloride Methylenbisthiocyanate (MBT) -Bromo--nitrostyrene Fatty amines 3) Others 23.400 21.400 71.152 13.200 18.000 12.150 56.800 56.400 9.500 17.200 10.000 19.800 400 11.000 6.900 2.400 20.000 4.412 2.250 20.800

64.100

Others

15.000

2.270 231

Total amount of non-oxidative biocides

Reference:

248.115

27.250

20.000

125.400

UK, F: IPPC Reference Document on the Application of BAT to Industrial Cooling Systems (11/2000) NL: IKSR Synthesebericht Antifoulings und Khlwassersysteme (Entwurf 15.11.2001) D: This study (F+E 200 24 33, Januar 2001) 1) All cooling water systems 2) Only recirculation cooling systems, NaOCl is indicated as Cl2 3) Consumption of fatty amines in a costal power plant

70 5.5 Elimination of chemicals in cooling systems and sewage plants Alongside the intended reaction, the greater part of the biocides and conditioners is eliminated in the cooling system through hydrolysis, adsorption, biological degradation or evaporation and is thus not relevant for surface waters. This is especially true for oxidatively acting biocides, whereby the in part more stable disinfectant side-products (including trihalomethane, AOX, and bromate) must still be considered. For the non-oxidatively acting biocides the elimination of fluorescent bacteria toxicity is used as the criterion for inactivation, whereby the elimination factors involved can usually not be distinguished. Organic corrosion inhibitors, dispersents and hardness stabilizers are usually not developed with ready biological degradability in mind, rather it is intended that they should be stable in the cooling system. The addition of readily degradable carbon sources in cooling systems would be counterproductive, since these serve as a substrate for the formation of biomass. Nonetheless, poorly or non-degradable compounds undergo some degree of elimination in cooling systems. Thus, for the phosphonic acid 2-phosphono-1,2,4-butantricarbonic acid (PBTC), for example, it is reported that in the presence of oxidative biocides (chlorine, bromine, ozone) as well as through photolysis an average of 20% of the PBTC is degraded after its addition in cooling systems (BUA 1996). For indirectly discharged cooling waters the elimination ability of the municipal sewage plant must also be taken into account, for which some examples are given in Sect. 6.4.2. In the Annex the substance-specific data are presented for hydrolysis, adsorption tendency (water solubility, log Pow) as well as for the elimination or biological degradability. From this information, the behavior in cooling towers and in municipal sewage plants can be estimated, at least roughly. 5.6 Regulatory control of cooling water discharges The discharge of cooling water is regarded as wastewater emission and accordingly regulated through the still valid Annex 31 to the Framework-wastewateradministrative procedures, which was revised in 2002 and is to be included again as Annex 31 in the German Wastewater Ordinance (Anonymous 2001). In the course of surveying the companies (cf., sect. 5.3) the responsible authorities were asked to provide information on the usual practices in controlling cooling water discharges. Of

71 special interest here was the experience with the fluorescent bacteria test, since it has been used since 1994 as the legally established standard test for determining the outflow shut-off pauses for cooling systems with non-oxidative biocides. The answers of the representatives of the agencies (as a rule, the local water authorities) were very non-uniform. In some districts only direct dischargers were considered, while in others all plants above a certain flow-volume were considered. Data on fluorescent bacteria toxicity were often only requested for direct dischargers. A systematic presentation of the experiences and results is not available. In personal discussions with representatives of the agencies it was reported to us that in general no significant pollutant loads were expected from cooling water emissions. Moreover, cooling water discharges were often sampled together with partial flows from other parts of the commercial operations, so that a separate compilation of the data for cooling systems would require a rather substantial effort. In our questionnaire distributed to the plants the persons responsible were also asked about the measured wastewater parameters (governmental agency control or their own). Of 156 plants with open circulation cooling or flow-through cooling that provided the corresponding data, for only 23% were data on fluorescent bacteria toxicity given. Some based their answers on single measurements taken several years previously, which were not regularly repeated. And even when only those cooling systems were considered for which biocides were known to have been used, a similar picture is seen. For only 1/3 of these cooling plants (n=52) were fluorescent bacteria-test results available. In 2/3 of the cooling systems with biocide use no fluorescent bacteria tests were performed, even though 40% of these discharged directly. Only in isolated cases was it mentioned that elimination curves for the biocides used were even on hand. In summary, it can be concluded from these data that there is a notable deficit in the performance of fluorescent bacteria tests for the evaluation of cooling water discharges. 5.7 Literature and database research The results of the research on the environmentally relevant substance properties of the cooling water chemicals are presented in the Annex. Along with the literatureand database research the information provided by the producers is documented as

72 personal communications. As a rule, the data provided by the manufacturers could not be verified from original data. The comprehensive evaluation of the ecotoxicity and degradability is presented in Table 20. As expected, the aquatic ecotoxicity of the biocides is quite high, while the corrosion inhibitors and hardness stabilizers can be described as relatively "nontoxic". With the exception of glutaraldehyde a ready degradability has not been shown for any of the organic biocides or conditioners. Only for quarternary ammonium compounds was a complete degradation reported in several studies (without respecting the 10-day window). In some cases there are still some significant data gaps. For some of the applied triazole compounds (e.g., butylbenzotriazole, chlorotolyltriazole) and phosphonic acids

(hydroxyphosphonoacetic acid, tetraalkylphosphonium chloride) there are virtually no data available. For PBTC and ATMP, in contrast, there are extensive substance dossiers (UNEP 1996, UNEP year not indicated). Also, little is known about several biocides such as the BCDMH-related 1,3-dichloro-5-ethyl-5-methylhydantoin or the QAV dimethylcocobenzylammonium chloride.

73 Table 20: Summarized evaluation of the eotoxicity and degradability of cooling water chemicals
Daphnia Biocides Chlorine Sodium hypochlorite BCDMH Ozone Hydrogen peroxide Glutardialdehyde Org. Bromo compounds Isothiazolines Org. sufluric compounds QAV 3 2 3 3 2 2 2-3 3 2-3 3 3 2-3 3 3 1 2 1-3 3 2-3 1-2 1-3 1-2 2 3 3 +++ + + + +-++ 3 1-0 + (DMH) Fish Bacteria Biol. Degradation

Corrosion inhibitors and Hardness stabilizers Phosphonic acids Triazoles Polycarboxylates Triethanolamine Zinc- /copper salts Molybdate Silicate 0 0 0 0 3 0 0 0 1 0 0-1 2-3 0 0 0 0 0 0 0 1 + + + +

The classification of ecotoxicity was performed following the criteria of the GefahrstoffVO: Classification number 0 1 2 3 Classification Non-hazardous Hazardous Toxic Very toxic LC50 or EC50 >100 mg/l 10-100 mg/l 1-10 mg/l <1 mg/l

The classification of bio-degradability was made following the criteria: +++ ++ + Readily bio-degradability proven Bio-degradable without 10 days window Poorly bio-degradable or inconsistent results

74

6 Evaluation
6.1 Composition of cooling water As a rule no ecotoxicity could be found for the cooling waters investigated. After a shock treatment with biocides a higher toxicity was measured, but this was completely eliminated within the prescribed inactivation period. Only in one plant with continuous addition of biocides (BCDMH) was a high toxicity found after an additional batch treatment with isothiazolinones. Although the outflow was not closed during treatment in this case, an indirect discharger was involved, so that no direct hazard for the surface waters was anticipated. In Germany the first results on the ecotoxicity of cooling water samples treated with biocides were presented by Brnert and Hagendorf (1992). For one indirect discharger the effectiveness of a not further specified biocide was inactivated after 4 h, but for another, also after 4 h, a G L- or GA-value of up to 192 was still measured (Brnert and Hagendorf 1992). Gellert and Stommel (1995) investigated the cooling water from 41 production plants and found among plants with flow-through cooling none exceeding the legally prescribed threshhold valules for fluorescent bacteria toxicity (GL12). For plants with circulation cooling with sodium hypochlorite shock treatments the fluorescent bacteria toxicity correlated with the chlorine concentration (up to 2.75 mg/l). The highest toxicity was measured immediately after the treatment of a circulation water with methylenebisthiocyanate (G L=1024). After 5 days the fluorescent bacteria toxicity of this biocide was also completely eliminated (Gellert and Stommel 1995). In the most extensive data collection on the ecotoxicity of cooling water up to that point, Diehl and Hagendorf (1998) presented 248 test results. The medians for the Daphnia, fluorescent bacteria and algae toxicity lay at GD=6, GL=1 and GA=6, the maxima at G D=128, GL=16 and GA=32. In comparison to the other biotests the fluorescent bacteria test was found to be less sensitive (Diehl and Hagendorf 1998). In a Dutch study the cooling water from 14 plants was investigated and a clear relationship was demonstrated between the substance concentration and the ecotoxicity of cooling water samples. In general, cooling waters treated with the non-oxidative biocides isothiazolinone, DBNPA, -bromonitrostyrene and methylenebisthiocyanate showed higher toxicity (up to a dilution

75 factor of 769) than those treated with the oxidative biocides sodium hypochlorite and ozone (< dilution factor of 2; cf., Baltus et al. 1999). 6.2 Emission route for cooling water chemicals The most important emission route for cooling water chemicals or their reactionproducts (e.g., AOX) is the outflow, or for indirect dischargers the municipal sewage plant. In open cooling circulations a portion of the oxidative biocides can be stripped out in the cooling tower (cf., sect. 6.4.1 for the example of chlorine, chlorine dioxide). In addition, chemicals dissolved in water are emitted in the form of expelled water droplets (ca. 0.01% of the water flow). Along with the chemicals added for conditioning, these also contain salts and microorganisms from the cooling waters (VDI 3784: 1986). 6.3 Elimination behavior of cooling water biocides Having information on the elimination time, for which period the outflow is to remain closed after a shock treatment with a non-oxidative biocide, provides an opportunity, according to Annex 31 of the AbwV regulations, to limit the effort involved in direct on-site measurements of fluorescent bacteria toxicity. This makes it all the more important to specify precisely the conditions for performing these tests. The results of the present study show that predictions about the decline in the biocidal effects should not be considered separately from the inoculum concentration. At a lower activated-sludge concentration (30 mg d.s./l) the fluorescent bacteria inhibition from isothiazolinone and Bronopol was not completely eliminated even after 7 days, while at the highest concentration (1 g d.s./l) a decrease within 1 day to G L<12 was observed; here elimination through adsorption played an important role. According to statements from cooling tower operators and manufacturers of cooling water conditioners, a closure of the outflow for more than 24 h is usually not practical. In the publication from German et al. (1996) 4 elimination curves for cooling water biocides in the fluorescent bacteria test are presented. The static experiments were inoculated with activated sludge from a municipal sewage plant (0.5 g d.s./l), and the determination of the fluorescent bacteria inhibition was made at a dilution of 12, as in the present study. Methylenebisthiocyanate (20 mg/l) showed an inhibitory effect of ca. 80%, which was also not completely eliminated after 24 h. In contrast, a

76 quarternary ammonium compound, isothiazolinone, and a

dibromonitrilopropionamide-formulation (at product-concentrations of 10, 70 and 100 mg/l) showed inhibitory effects even at the start of only 27% to 50%, so that under these experimental conditions no effective biocidal effect was detected even by using individual species such as the fluorescent bacteria. After 2 to 3 h, all 3 compounds were below the threshold value of 20% inhibitory effect (German et al. 1996). In a Dutch study the decrease in the fluorescent bacteria inhibition by cooling water biocides was investigated specifically with regard to hydrolysis as an elimination factor. Here the surface water used for cooling was tested, without any additional inoculum, after removing zooplankton by filtration through 25 m pores. Within 96 h the fluorescent bacteria inhibition from DBNPA (12.5 mg/l), -bromo--nitrostyrene (2.5 mg/l) and -bromo--nitrostyrene/methylenebisthiocyanate (4 and 2.2 mg/l) was completely inactivated, while for isothiazolinone (1.5 mg/l) and glutardialdehyde (50 mg/l) only minimal elimination was observed (Baltus et al. 1999). Gartiser and Scharmann (1993) utilized unconditioned cooling water without addition of any further inoculum to determine the inactivation kinetics of 3 biocides at their usage concentration. For the biocide based on isothiazolinone the G L-value indeed decreased within 13 days from G L=128 to GL=16., but did not reach the legally prescribed threshhold value G L=12. The fluorescent bacteria toxicity of the biocide based on hydrogen peroxide was completely degraded in 3-6 days. A dye used against algal growth was already below the threshhold value of GL=12 when tested at its application concentration (Gartiser and Scharmann 1993). The question of establishing the experimental conditions for the determination of "elimination curves" for cooling water still remains open. Elimination curves obtained under the experimental conditions of the VCI-working group "Biocides in cooling circulations" with a higher inoculum concentration (activated sludge, 500 mg d.s./l), favor elimination through adsorption and correspond to the design of an inherent degradation test. Comparably high biomass concentrations are not normally found in cooling circulations. The biomass in water channeling systems is sessile, primarily in the form of biofilms; the microbe count in the water is 10 3 to 104 fold lower. Biofilms in water-channelling systems seldom exceed a thickness of more than 1 mm and contain 98% to 99% water. The greater part of their dry mass consists of extracellular polysaccharides, and only maximally 10% of the dry mass is present in

77 the form of cells (Cloete and Brzel 1992). A comparative measurement of the elimination behavior of cooling water biocides in a laboratory test and in real plants is important, in order to interpret the data correctly. However, here there are scarcely any published results available. For all such measurements at real plants it is important that the decrease in the fluorescent bacteria toxicity is not caused by dilution with added make-up water. Therefore, the outflow must be closed during such an experimental series. When additional information on the biological degradability is called for, especially when evaluating direct dischargers, results on ready biological degradability and/or elimination curves with lower inoculum concentrations (e.g., 30 mg d.s./l in conformance witih the conditions for the OECD 301 Test of "ready bio-degradability") should be called for. 6.4 Choice of active substances 6.4.1 Biocides The choice of a biocide depends in part on the organisms whose growth is to be controlled. Especially for preventing algal growth, copper sulfate, quarternary ammonium compounds and photosynthesis inhibitors based on triazine derivatives are used. There are also some contradictory reports on experience concerning the effect spectrum of cooling water biocides. Thus, for example, it was reported that glutardialdehyde and quarternary ammonium compounds can only be effectively used against algae, while isothiazolinones showed the best effect against bacteria. With methylenebisthiocyanate biofilm formation could not be effectively prevented (Mattila-Sandholm and Wirtanen 1992). Mussels cannot be effectively combatted in cooling systems with chlorine-releasers, because they protect themslves by closing their shells. In these cases non-oxidative biocides such as QAV are used. In one investigation isothiazolinone but also QAV and thiocarbamate also proved ineffective against several amoebae isolated from cooling water thought to be causing the spread of pathogenic Legionella (host organism). On the contrary, a growthpromoting effect was even observed in some cases (Srikanth and Berk 1993). In the following, the most important biocides will be described.

78 Chlorine and sodium hypochlorite Active chlorine added in the form of chlorine and sodium hypochlorite is the most important oxidative biocide for cooling plants. For cooling plants that use sea water as a heat carrier, usually 0.5-1.5 mg/l chlorine is continuously added. In conformance with the reference document on the IPPC-directive, this is also BAT (Anonymous 2000). Because of the bromides contained in the saltwater, different disinfectant byproducts are produced here than those found for fresh-water chlorination. In a French study bromoform, dibromoacetonitrile, tribromophenol and dibromoacetic acid were identified as the main components (Allonier et al. 1999). The VGB-cooling water guideline recommends a level for continuous addition of 0.1 mg/l free chlorine (for brackish water) and 0.25 mg/l (for sea water), and for batchwise shock-treatment chlorination concentrations of 2-3 mg/l free chlorine (Fichte et al. 2000). At the usual pH-values > 8 in cooling circulation, the biocidal effect of chlorine in the water is reduced, because the dissociation equilibrium is shifted away from hypochlorous acid in favor of the less effective hypochlorite-ion (OCI -). In this case it is better to use the more weakly dissociating hypobromous acid (HOBr), which dominates over the only weakly microbiocidally acting hypobromite-ion (OBr -) up to pH 9. Hypobromous acid is, as a rule, generated on site from sodium bromide by adding sodium hypochlorite (NaOCl). With decreasing pH an increasing proportion of the hypochlorous acid is stripped out during cooling tower passage (ca. 30-40% at pH 6 and ca. 10% at pH 8.5, Holzwarth et al. 1984, Baltus and Berbee 1996); therefore, it is recommended to use chlorine-releasing biocides in open circulation cooling in spite of their lower effectiveness at pH-values above 8 (Anonymous 2000). Depending on the water composition (including DOCand ammonium

concentration), pH-value and contact time, the use of free halogens as biocides leads to the formation of disinfectant by-products such as trihalomethane, chlor- and bromamine as well as adsorbable organic halogen compounds (AOX).

79 Chlorine dioxide In the treatment of drinking water chlorine is in part replaced by chlorine dioxide, in order to minimize the formation of AOX, especially halogenmethanes. Chlorine dioxide reacts substantially more weakly with complex organic molecules and ammonium and thus forms less AOX. Ocasionally, chlorine dioxide is also used in the cooling water field, whereby it is usually produced on site through the reaction of chlorine gas with sodium chlorite (NaClO 2). Chlorine dioxide is considerably more volatile than hypochlorous acid and consequently can be more easily stripped out from open cooling plants (Anonymous 1991, Groshart and Balk 2000). BCDMH Among the organic chlorineand bromine-releasers, 1-bromo-3-chloro-5,5-

dimethylhydantoin (BCDMH) is preferentially used. Here the actual active compound, hypobromous or hypochlorous acid, is released after a time delay. BCDMH hydrolyzes rapidly and because of its limited water solubility it is usually added over a separate sidestream (over BCDMH filtered cooling water) (van Donk and Jenner 1996). The organic residue after hydrolysis, 5,5-dimethylhydantoin, is poorly bio-degradable, but has a weak tendency to adsorption (Anonymous 1994). It has been reproted that the biocidal effect of 1-bromo-3-chloro-5,5-dimethylhydantoin is 10 to 20 fold higher than that of chlorine-releasers (Mattila-Sandholm and Wirtanen 1992). Related compounds such as 1,3-dichloro-5,5-dimethylhydantoin or 1,3-dichloro-5-ethyl-5-methylhydantoin are also occasionally used. Ozone Ozone is a highly effective oxidative biocide. While ozone holds a firm position in the treatment of swimming-pool water, there is little practical epxerience available on the use of ozone in cooling water treatment (Schmittecker et al. 1997, Schmittecker et al. 1999, Wasel-Nielen and Baresel 1997). In the data we compiled on cooling water practices (cf., sect. 5.3), there was only one plant that used ozone. In a study of the Landesanstalt for Environmentschutz Baden-Wrttemberg the following advantages of treating cooling water with ozone, as compared to chlorine and non-oxidative biocides, were listed: effective inhibition of biogrowth, no storage necessary since

80 ozone is produced on site, no formation of AOX, attainment of the CSB-threshhold, elevation of the ozone concentration in the air near the ground by maximally 1 g/m 3 (Schweitzer 1995). The chemical reactions of ozone in water have been described by Rice and Wilkes (1992). It is reported that ozone decomposes at pH-values above 8 to free hydroxyl radicals. These have a stronger oxidizing effect than ozone itself, but at the same time they also have a very short half-life of a few micro seconds. Especially with high concentrations of hydrogen carbonate as a radical-trap, the disinfectant effect of ozone at pH-values >9 is poor. In the ozonation of bromidecontaining water, the bromide is oxidized to hypobromite-ion (OBr -), which is rapidly hydrolyzed to the equally biocidal hypobromous acid (HOBr) or to the genotoxic bromate (BrO3-) ion. As disinfectant by-products compounds such as bromate, bromoform or monobromoamine are formed, some of which show a genotoxic effect (Rice and Wilkes 1992, Gunten and Hoign 1994). Comparative measurements with the Ames test on the mutagenicity of disinfectant-by-products of ozone, chlorine and chlorine dioxide in the treatment of drinking water, showed a substantially higher mutagenicity after disinfection with chlorine. At a dose of 3 mg/l ozone a slight increase in the mutation rate was found, but at higher ozone concentrations (10 mg/l) this was eliminated (Kool and Hrubec 1986). Usually, ozone is added to the cooling water continuously in very low concentrations of 0.1 to 0.3 mg/l (Wasel-Nielen and Baresel 1997)(Viera et al. 1999). Production occurs directly on site using high voltage, with an energy expenditure of ca. 10 kWh per kg ozone. Thus, for cooling systems in energy generation, up to 0.1% of the energy generated must be used for the production of ozone (Anonymous 2000). Due to the high oxidative potential of ozone decomposition, when used at high concentrations, the intermediate hydroxyl radicals formed may attack structures made of plastic or wood. Organic conditioners, for example dispersents, can also be degraded by ozone. In order to keep the ozone content low, the cooling water should have the lowest possible content of organic material.

81 Hydrogen peroxide In comparison to the other oxidative biocides, hydrogen peroxide is only effective at higher concentrations (>15 mg/l; cf., van Donk and Jenner 1996). Reports of lower effective concentrations were questioned on the part of industry (cf., table 9). Due to iron- or copper-catalyzed decomposition, it only has a very short half-life. In some cases, the effectiveness of hydrogen peroxide was improved through the use of special catalysts. Rarely, peracetic acid is used as an organic oxygen-releaser in the cooling field. This compound is readily biologically degradable, but under unfavorable conditions it is corrosive. Isothiazolinones One of the most important non-oxidative cooling water biocides consists of a mixture of compounds from the isothiazolinone substance group: 5-chloro-2-methyl-4isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one. They are already effective in the concentration range below 1 mg/l. The mechanism of action involves an inhibition of cellular proteins (Groshart & Balk, 2000). The isothiazolinones hydrolyze slowly (t1/2 = 7 d at 30C and pH 8) and are biologically not readily degradable. In experiments with laboratory sewage treatment plants a limited mineralization of the compounds (<25%) was demonstarted (Baltus & Berbee, 1996; Krzeminski et al., 1975). Unpublished test reports from Thor GmbH in simulation tests

(Water/Sediment-test according to BBA IV 5-1) confirm the rapid primary degradation of the investigated isothiazolinones, but only ca. 10% was mineralized to
14

CO2. Thus, 5-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-

one form stable metabolites, which are to a large part adsorbed in the sediment in the test system (Schoester 2001). Isothiazolinones are rapidly bound to biomass. Adsorption and bioaccumula tion in the biomass are major elimination factors. Investigations with the bacteria Pseudomonas aeruginosa and P. fluorescens showed that the concentration of 5chloro-2-methyl-4-isothiazol-3-one inside the cells was a factor of 400 higher than outside (Diehl and Chapman 1999).

82 Quarternary ammonium compounds Quarternary ammonium compounds (QAV) act by binding to the cell membrane and are also not readily biologically degradable. In sewage plant passage they are largely eliminated by adsorption to the activated sludge. With an adapted inoculum or after longer incubation times, an extensive mineralization of QAV has been reported. The most important representative in the cooling water field is alkyldimethylbenzylammonium chloride. Organic bromine compounds The non-oxidatively acting dibromonitrilopropionamide (DBNPA) is hydrolyzed rapidly to the in part still biocidal compounds dibromoacetonitrile, dibromoacetamide, monobromonitrilopropionamide and cyanoacetamide. Only substance-specific

degradation data are available (Blanchard et al., 1987). The substance itself is presumably poorly biologically degradable (Groshart & Balk, 2000). Other members of this group include 2-bromo-2-nitropropan-1,3-diol (Bronopol) and -bromo-nitrostyrene, but no degradation data on these compounds is available. Bronopol is hydrolyzed slowly (t1/2=14 d at 30C and pH 8), but in contrast -bromo-nitrostyrene hydroyzes rapidly (t 1/2=20 min at 30C and pH 8) (Baltus and Berbee 1996). Glutardialdehyde Glutardialdehyde is only used occasionally in the cooling water field. The mecha nism of action is based on the denaturation of proteins. Glutardialdehyde is less toxic for aquatic living organisms as compared to the other biocides, and the concentration used is correspondingly higher. The compound is readily bio-degradable. 6.4.2 Cooling water conditioners Phosphates Phosphates and polyphosphates have been used for a long time as corrosion inhhibitors. For very soft waters a cooling water treatment with zinc ions or inorganic phosphates is often used for steel as a satisfactory protection against corrosion. The concentration of orthophosphate used to inhibit corrosion is maximally 15 mg/l. This corresponds to a concentration of phosphorus of ca. 5 mg/l. In the opinion of several

83 of the manufacturing companies, in the present study the balance sheet loads of inorganic (poly)phosphates are probably underestimates. However, this could not be confirmed. Even if the consumption of phosphoric acid is assigned to the phosphates (as was the case), in comparison to the organic phosphonic acids substantially fewer plants report the use of inorganic phosphorus compounds. The evaluation of product information showed that phosphates were primarily used in combination with phosphonic acids and/or polycarboxylates, as well as with zinc or molybdates. This explains why the overall balance gave phosphate concentrations lower than would correspond to the concentration added if used alone. Thus, for example, a good protective effect was achieved for non-alloy steel in cooling circulations using sodium polyphosphate as well as sodium polyphosphate or polyacrylate in combination with zinc sulfate. The exclusive addition of zinc sulfate did not reduce the rate of corrosion (Naumann and Schimke 1991). Hence, the calculated concentrations (1.8 mg/l) also are lower than the values that would be necessary for the sole addition of phosphate (cf., table 16). Of note is the frequent use of inorganic phosphates in the water added to the heating vessel. Phosphonic acids Phosphonic acids are used in very sub-stoichiometric concentrations for hardnessstabilsation (treshold-effect) and in higher concentrations as corrosion inhibitors. They are biologically poorly degradable and stable against hydrolysis. In the concentration range of 5-20 mg/l some phosphonic acids have good properties as corrosion inhibitors for steel (all-organic operation). In conjunction with zinc, however, only one-half the phosphonic acid concentration is needed for a corrosion inhibiting effect. Under the phosphonic acids which could be unequivocally identified in the product information provided, the frequency distribution was HEDP >

hydroxyphosphonoacetic acid > PBTC >> ATMP and DTPMP. By and large, this agrees with the frequency of the surveyed phosphonic acids

(hydroxyphosphonoacetic acid > HEDP >> ATMP = PBTC = DTPMP). However, owing to the poor documentation, the predominantly consumption numbers for phosphonic acids could not be assigned to particular substances.

84 The elimination of PBTC in sewage plants appears to occur, according to laboratory experiments, 0-60% through adsorption, as well as 60-100% through flocculation (Pelimination (BUA) 1996). Metzner et al. measured the elimination capacity for HEDP and ATMP in municipal sewage plants to be ca. 50% (Metzner 1990). The elimination of some other phosphonates in municipal sewage plants is reported to be on the order of 50%. Adsorption experiments with radioactively labeled phosphonic acids at environmentally realistic concentrations showed that 23-70% of EDTMP and 80-90% of DTPMP was adsorbed to various activated sludges (Gledhill and Feijtel 1992). In general, phosphonates influenced the phosphate-precipitation, as the third purification step, since they form complexes with iron(III)-ions. However, this influence is compensated by the higher concentrations of flocculants (Horstmann and Grohmann 1986). The dependence on pH-value and the concentration of heavy metals for the adsorption of phosphonic acids onto clay minerals was investigated by Nowack et al. (1999). They found that the adsorption decreased substantially above pH 8. While copper and iron ions only had a limited influence on adsorption, it increased noticeably in the presence of calcium (Nowack and Stone 1999). In the outflow of 7 municipal sewage plants with phosphate-precipitation, phosphonate concentrations were generally below the detectability limit (ca. 2 g/l). The elimination in the sewage plant was over 80% (Nowack 1998). Only a few studies considered the behavior of the metabolites of phosphonic acids. In a recent Dutch study the in- and outflow-concentrations of aminomethylenephosphonic acid (AMPA), a degradation product of ATMP, were determined in 5 municipal sewage plants. In some cases, higher concentrations were measured in the outflow. (Meerkerk and Puijker 1997). The proportion of the cooling water conditioners (for direct dischargers) in the total load of AMPA in a part of the drainage basin of the Maas River was estimated at 28%. The contribution by municipal sewage plants was in the same range. An additional source for AMPA is the degradation product of the herbicide Glyphosat (Meerkerk and Puijker 1997). Nowack et al. (2001) investigated 2 diphosphonates which are formed in the catalytic degradation of ATMP in the presence of manganese(II)-ions and molecular oxygen. These compounds were also effectively eliminated in the sewage plant (>80%) (Nowack 2001).

85 In the surface water the photodegradation of phosphonic acids surpasses bio logical degradation by far, especially when they are present in the form of iron complexes (Gledhill and Feijtel 1992, Steber and Hater 1997). It should be noted, however, that only the primary degradation of the substances was determined. Several studies document a substantial photodegradation of phosphonic acids in the presence of equimolar amounts of iron ions (ca. 44% to 82% at pH 7 in 17 days with natural sunlight illumination; Gledhill and Feijtel 1992). Unpublished data of the Bayer AG on the photodegradability of PBTC, HEDP and ATMP at a test concentration of 10 mg/l in the presence of 0.2 mg/l iron-ions indicate a PBTC-degradation of 53-81% within 24 h, but a negligible degradation of HEDP and ATMP over this time. Without the addition of the iron-ions no photodegradation of PBTC was detected. The presence of nitrate also favored photodegradation. Adsorption experiments with radioactively labeled PBTC (1 mg/l) showed that 95% adsorbed to the activated sludge (Kleinstck 2001). However, the photolysis of the iron(III) complexes of ATMP and DTPMP, unlike that of the iron(III)complex of EDTA does not lead to biologically degradable compounds (Nowack and Baumann 1998). As an additional, relevant degradation mechanism in natural surface waters, the catalytic oxidation in the presence of manganese-ions and oxygen has been discussed (Nowack and Stone 2000). The analytical determination of phosphonic acids and their metabolites in surface waters has so far been limited by the relatively high detection threshold of ca. 2-13 g/l with the method of Klinger et al. (1999) and Nowack (2001). However, since it can be expected that phosphonic acids will be found in surface waters, their elimination in the treatment of drinking water has been investigated. The results show that all phosphonates are removed by precipitation with iron(III)- or aluminumsalts under practically relevant conditions. Using ozone only the nitrogeneous phosphonic acids are completely converted, whereby the herbicide Glyphosat and its main degradation product aminomethylphosphonic acid (AMPA) are formed as oxidation products (Klinger et al. 1998). Polycarboxylates Polycarboxylates are generally regarded to include the water-soluble negatively charged polymers (polyelectrolytes) which are primarily used (alongside zeolites) as phosphate substitutes in the wash-detergent field. These are homopolymers of

86 acrylic acid as well as co-polymers of acrylic acid and maleic acid with a mean molecular weight (MW) of ca. 70,000 (Opgenorth 1992). These are not classified as defined compounds with their own CAS-numbers, so no direct substance searches could be carried out, and reference had to be made to more general studies. This substance group is consequently not included in the Annex. In Germany ca. 20,000 t/a polycarboxylates are used in wash detergents (Kaiser et al. 1998). In the cooling water field, hardness stabilizers and dispersants of polyacrylates, polymethacrylates and polymaleinates, sulfonated copolymers and polyacrylamides are all grouped together under polycarboxylates (Piegsa 2001b). Polycarboxylates with MW of 500-20,000 are used in the cooling water field for hardness stabilizing, and higher molecular forms (up to MW 50,000) as dispersents (Andres et al. 1980). In the product information these compounds are often listed under polycarbonic acids. All these compounds are poorly bio-degradable, whereby a low molecular weight favors degradation. Thus, the copolymer of acrylic acid and maleic acid at a MW of 70,000 is only degraded 20% after 30-90 days, and at a MW of 1000 up to 45% (Opgenorth 1992). In municipal sewage plant polycarboxylates are primarily eliminated through precipitation with calcium and magnesium hardness and adsorption to the activated sludge, whereby in laboratory experiments (ZahnWellens-test, SCAS-test, laboratory sewage plant) values up to 98% were determined for higher molecular weight compounds. Below a MW of 10,000 the adsorption tendency of a copolymer of acrylic acid and maleic acid decreases substantially (Opgenorth 1992). In a recent study on the degradability of 2 radioactively labeled polycarboxylates based on acrylic and methacrylic acid and styrol at a MW of 55,000 and 7,000 only negligible mineralization in the CO 2evolution-test (OECD 301 B) was measured. In the SCAS-test (semi-continuous activated sludge test according to OECD 302 A) < 1% of the polycarboxylates was in the aqueous phase, and the bulk (96-99%) was adsorbed to the activated sludge (Jop et al. 1997). The aquatic ecotoxicity of polycarboxylates is generally low. The EC 50 for Daphnia, and fish was > 200 mg/l for the copolymer of acrylic acid and maleic acid with MW= 4500 or 70,000 (Kaiser et al. 1998, Opgenorth 1992). In the activated sludgerespiration-inhibition test (OECD 209) up to a concentration of 3,200 mg/l no significant inhibitory effect against the activated sludge could be detected (Jop et al.

87 1997). Similar data were reported by Guiney et al. (1996). Also the polycarboxylates used in the wash-detergent field have a low ecotoxicity. A potential to remobilize heavy metals and for bioaccumulation was not observed (Schberl and Huber 1988). Dispersants based on acrylic acid with a MW of 45,000 were weakly ecotoxic in the activated sludge respiration-inhibition test, fluorescent bacteria test and algae test with EC50 values <100 mg/l. However, for aquatic crustaceans of the species Ceriodaphnia dubia chronic effects due to physical damage have been reported, the NOEC (no effect concentration) here was 1 mg/l (Guiney et al. 1996). Triazoles Triazoles are predominantly used as corrosion inhibitors for copper and its alloys. The most important compounds are benzotriazole and tolyltriazole (Andres et al. 1980). The triazoles generally have low aquatic ecotoxicity. Few data are available on the degradability or elimination, but triazoles are generally regarded as not readily degradable. The degradability of tolyltriazoles appears to depend on the position of the methyl group. While 5-methyl-1H-benzotriazole reported to be degraded in cooling towers, 4-methyl-1H-benzotriazole is stable (Cornell et al. 2000). These data agree with the Zahn-Wellens test results (96% elimination for 5-methyl-, 0% for 4methyl-1H-benzotriazole; cf., annex). For other compounds there are serious data gaps as regards the ecotoxicity and eliminability. Heavy metals Zinc and molybdenum are also used as corrosion inhibitors in open industrial cooling circulations. When these heavy metals are indirectly discharged, elimination in the sewage plant through adsorption should also be taken into account. It is well known that adsorption to the activated sludge occurs, and consequently threshold limits have been established according to the sewage-sludge regulations (for removal for spreading on agriculturally used soils, for zinc 200 mg/kg). Zinc According to Annex 31, zinc salts may not be used in open cooling systems of power plants. The evaluation of the questionnaires confirmed that this restriction was respected. In industrial cooling systems zinc is often used in combination with phosphates or phosphonates. In combination with phosphate the pH in the

88 circulation water is maintained in the neutral range through the addition of mineral acids, in order to avoid excessive deposition of poorly soluble zinc hydroxides or calcium phosphate. In contrast, when using phosphonates the addition of acid is not necessary (Piegsa 2001b). A balance for the zinc loads in the in- and outflows of municipal sewage plants revealed that ca. 50% was eliminated with the excess sludge and 50% passed with the wastewater in the final settling tank (pond) (Kroiss and Zessner 2001). Zinc is an essential trace element for animals and plants, but in higher concentrations is toxic. The zinc content in the outflow of industrial cooling circulations is limited to 4 mg/l according to Annex 31. Molybdenum Several manufacturers remarked that molybdate is primarily used in closed systems. Nonetheless, its use in open systems is documented for 6 plants (from the foodstuffs industry, automobile industry, chemical industry and one steel plant), so its loads may not be neglected. The use in open systems also agrees with the producers information for several products. Molybdenum is an essential trace element for plants and is contained in the enzyme for nitrogen fixation (nitrogenase), as well as in several animal enzymes. Consequently, small amounts of molybdates are added to some fertilizers. However, higher concentrations can be toxic for both plants and grazing animals (Parker 1986). In comparison to zinc salts, the aquatic ecotoxicity of molybdates is substantially lower. Investigations on the adsorption of molybdates to activated sludge are not available, but it can be supposed that there is an enrichment on the activated sludge.

89

7 Recommendations
Cooling plants are components of complex thermal technical processes and furthermore are dependent on the available water and the local conditions. Consequently, an optimization of the cooling and the selection of the conditioners should always take into account the whole process. 7.1 Energy conservation measures Cooling plants are used to remove from the industrial process excess heat produced during the consumption of primary energy. Hence, a rational use of the energy leads not only to cost savings, it also reduces the emissions of greenhouse-gases and the effort required of the cooling plants including the amount of chemicals. In optimizing the process internal measures to save energy should thus have the highest priority. In addition, the use of the industrial waste heat for thermal energy is by far the most environmentally friendly system for the provision of low-temperature heating. For this reason, an increased development of thermal energy production with the utilization of industrial excess heat is being sponsored (Roth et al. 1996). 7.2 Technical solutions In the reference document "Industrial cooling systems" the various technical possibilities for the optimization of cooling plants are comprehensively presented, so here only the essential elements need to be mentioned. The choice of the cooling system depends on the local givens, such as, e.g., the capacity of the surface water to be used. From the questionnaire to the cooling plants it was surprisingly learned that, in 34 of the 110 surveyed plants with open circulation cooling, ground water was used for cooling purposes and in 12 plants drinking water. Indeed, in 3 plants more than 100,000 m 3 drinking water were used annually for cooling. Thus, the water supply can be a significant cost factor and influence the site selection. Regarding the choice of basic materials, it seems that because of the complexity of cooling water systems there is no ideal basic material. Nonetheless, the choice must be made according to the specific requirements. In the VGB-cooling water guideline recommendations for basic materials for pipes, heat exchangers and water

90 chambers are provided with considerations of the water composition (chloride content). From the experiences reported by the employees of the plants included in this study, it became clear that already through simple measures the cost intensive use of cooling water biocides could be reduced, in some cases by 90%. Thus, for example, the orientation of the cooling tower with respect to the sun plays a decisive role for the growth of algae, the primary producer of biomass. Already through the placement of the cooling tower and/or its subsequent shading, algal growth can be substantially reduced. Pre-treatment of cooling water: Through treatment of the added water dispersed material can be removed by filtration procedures, or the carbonate hardness and/or the salt content of the added water reduced with an ion-exchanger. Thereby, the amount chemicals required can be reduced and higher concentration cycle factors can be adjusted downward. However, in making a comparative evaluation, the chemicals required for the decarbonization and regeneration of the ion-exchanger must also be taken into account. Thus, the regeneration requirements for hydrochloric acid and sodium hydroxide fluctuate between 110% and 250% of the stochiometrically called for amounts, and for sodium chloride solution (NaCl) between 150% and 300% (Greiner et al. 1987, Kohler 2000). For power plants, because of the large demand for water, the treatment of the added water is generally limited to the primary interfering substances (mechanical prurification by filtration, decarbonation with sulfuric acid or calcium hydrate) (Fichte et al. 2000). Side-stream filtration: The addition of chemicals in open circulation cooling can also be reduced by continuously cleaning a partial-stream using filter technology. Thus concepts were developed to reduce the potential for biofouling in cooling systems by removing the nutrient materials. By introducing a reverse-flow washable sand filter upstream, the formation of the biofilm can be specifically limited to an easily controllable system and thereby protecting the downstream cooling plant. The biologically degradable carbon could be reduced in one plant with this technique to 1/4 of the original value. The thickness of the biofilm was corresponding reduced by 90% (Griebe and Flemming 1996). By cleaning just 1-3% of the circulating water over a continuously operating sand filter the consumption of sodium hypochlorite was reduced by 70-80% (Daamen et al. 2000). The VGB-cooling water guideline

91 recommends where appropriate the filtration of a side-stream of up to 5% of the total flow through the use of gravel-filters (Fichte et al. 2000). According to Kunz (1993) water circulations should be regarded as bioreactors in which growth should not take place". Based on the observation that many organisms protect themselves from colonization with bacteria by keeping their milieu iron-free, his biotechnological concept for limiting the biomass growth consists of seeking to specifically reduce the concentration of iron ions in the cooling system (Kunz 1993). Water-flow velocity: Periods of standing, stagnation phases and too low flow velocities are to be avoided, in order to minimize the risk of biomass formation, corrosion and scaling. In the reference document of the EU on industrial cooling systems current velocities of 1.8 m/s are described as a BAT-measure to protect condensers and heat-exchangers (Anonymous 2000). On the other hand, excessively high flow should also be avoided to prevent possible erosion corrosion phenomena and the formation of a protective layer. In the VGB-cooling water guideline a current-flow velocity range of 0.5-3 m/s is recommended (Fichte et al. 2000). Chemical-free operation: Occasionally, there are reports of attempts to introduce physical procedures for cooling water treatment. Thus, for example, in laboratory experiments electrical impulses were successfully used to control salt-water polyps (Hydrozoa arge) (Abou-Ghazala and Schoenbach 2000). Brief treatments with heat or microwaves are supposed to control the biology in cooling systems (Anonymous 2001c). However, corresponding descriptions of practical experience here are not available. 7.3 Process operation In the course of our visits to the plants, various employees reported to us that the amounts of chemicals to be added could be substantially reduced by regular cleaning of the cooling pans and sprinkling structures. It is known that biofilms are substantially more resistant to biocides than planktonic microorganisms. Moreover, killing the organisms present does not remove the biofilm itself, so that the possibility of a rapid recolonization and a continuation of the biofouling exists. Therefore, attempts are made to remove the dead biofilm with dispersants, but often this only partially succeeds (MacDonald et al. 2000). Insofar as the formation of biofilms with

92 acceptable thicknesses cannot be prevented in advance, mechanical cleaning methods are more effective. The VGB-cooling water guideline recommends the continuous cleaning of the cooling pipes with beads or brushes. A number of beadcarriers (sponge, granulate or polishing beads), coated with various materials (pumice, plastic, natural rubber, corundum), are available on the market (Fichte et al. 2000). There are also self-cleaning heat-exchangers available on the market in which particles 1 to 3 mm in diameter are carried with the cooling water current and before emission are reclaimed, e.g., via hydrocyclones (internal circulation, e.g. Klarex Technology B.V., the Netherlands). Fully automatic systems for the cleaning of condensers with compressed air have been developed. Larger surfaces are cleaned with high-pressure water streams (50-200 bar), and for pipes automatic cleaning systems are offered (brush-cleaning, sponge-bead-cleaning, swab-cleaning) (Paikert 1986). Closing the blow-down after a batch treatment with biocides until their demonstrated inactivation is an important instrument for protecting the sewage plant and the final tanks. Another advantage of closing the outflow is that the contact time of the biocides with the cooling water is increased and at the same time also their effectiveness. Carefully measuring the added cooling water chemicals to match the amount actually needed can potentially bring a great savings. For example, quick methods were developed for surveillance (monitoring) of the microbiological activity of cooling water samples. One possible measureable endpoint for this is the content of adenosine triphosphate (ATP) in the cooling water, which can be measured with luciferin as released light (Marczewski et al. 1995, Hild 2000). In addition, fluorescent bacteria toxicity is also used to monitor the effect of a present biocide. At G L-values >2 there was a distinct reduction of the total number of colonies (Schaaf no year given). Several companies offer measuring instruments for the determination of the fluorescent bacteria toxicity (e.g., Dr. Lange Lumismini, Merck ToxAlert , FHFurtwangen, BetzDearborn Bioscan TM). However, based on previous experience, these techniques are hardly ever applied in practice. Instead, a continuous conductance measurement in the outflow of open circulation cooling plants is state of the art in order to permit adjustment of the blow-down and with it the concentration

93 factor. Moreover, the dosing of the addition of chemicals can be guided by online measurements. Phosphonic acids can also measured with online phosphate analyzers (Elfering and Schmitz 1995). 7.4 Evaluation and selection of cooling water chemicals There are essentially 4 approaches with which the hazard potential from cooling water chemicals can be evaluated and applied as an instrument for the selection of an active ingredient. In the reference document of the European Commission 3 of these are presented in detail, so here only the basic outlines and possible associations will be mentioned (Anonymous 2000). 7.4.1 Water risk classes, the VCI-concept for open cooling systems The concept of the Verband der Chemischen Industrie was not developed for the evaluation of cooling water chemicals, rather for taking measures during possible leakages with water-hazardous substances. It proposes measures for monitoring and adjusting fresh-water and open circulation operation, as well as alternatives to oncethrough cooling, depending on the water-risk presented by the substances which could be introduced by leakages of cooling water. The basis is a derivation of waterhazard classes in conformance with European laws on hazardous substances from the Risk- or Gefahren-Phrases (R-Phrases), which are presented in the Verwaltungsvorschrift wassergefhrdender Stoffe" from 17 May 1999 (VwVwS) and expounded by the Commission for the evaluation of water-endangering substances (Kommission zur Bewertung watergefhrdender Stoffe 1999). Here each R-Phrase relevant for the protectable natural resource aquatic environment in combination with the protectable natural resources human health and soil protection is assigned a certain point value. The points for all the R-Phrases given to a particular substance are added up to give a total number of points. Here data gaps on ecotoxicity, degradability/bioaccumulation or acute oral/dermal mammalian toxicity are

compensated with additional default assigned values. With this total point value the water-risk class of a substance in accordance with the VwVwS can be determined. For prepared mixtures a water-risk class for the product is then generally determined from the consideration-limits (the sum of proportional values for each of the WGKsubstances present, scaled on a percentage basis corresponding to its proportion in the mixture) (Annex 4 VwVwS).

94 The VCI-concept links the total point value with the necessary cooling water safety measures (Anonymous 2001b, Anonymous 2000, Piegsa 2001b). For substances with a total point value > 5 (corresponding to WGK 2) depending on the contingencies of each individual case, measures for the monitoring of once-through cooling systems should be immediately undertaken and where appropriate the cooling water should be directed into wastewater treatment facilities (usually municipal sewage plants). However, the concept is only valid for direct dischargers. Although the VCI-concept is not intended for the selection of cooling water chemicals based on environmental aspects, the emission based approach for considering water-risk classes could be a possible instrument for a comparative evaluation of cooling water chemicals. However, since the assignment of water-risk classes especially considers the accident situation, the concept requires further development, e.g., in order to relativize the high ecotoxicity of a chemical to a known rapid elimination in cooling systems or sewage plants. In addition, for mixtures the consideration-limits (the sum of proportional values for each of the WGK-substances present based on the composition of the mixture) must be incorporated for the determination of WGKclasses. 7.4.2 "Benchmarking"-concept Benchmarking is a continuous process, in which a defining evaluation parameter is compared with a reference situation (benchmark). Here data from comparable situations are collected and compared with the target data of the reference situation in the form of a relative "ranking" so that every participant can recognize "where he stands", i.e., how far he is from the benchmark. In the reference-document of the EU "Industrial cooling systems" this instrument was applied to the procedure of risk evaluation for chemicals usually followed in the EU, as described in the Technical Guidance Document (Commission 1996). This is based on the ratio of the environmental concentration to be expected (predicted environmental concentration, PEC) to the threshold-concentration, below which no effects would be expected (predicted no effect concentration, PNEC). Here the PEC is calculated from emission scenarios from the discharged amounts, the elimination factors and the dilution; the PNEC is generally determined from laboratory ecotoxicity tests assuming certain

95 safety factors (up to a factor of 1000). In special cases the PNEC is also extended using model ecosystems as well as field studies. In Annex V of the WaterFramework-Directive of the EU (2000/60/EG) the procedure for obtaining the PNEC is to be introduced in order to define the chemical quality standards to be established by the member states. Here data on persistence and bioaccumulation are to be considered (environmental quality standards, EQS). The "benchmarking"-concept only envisages a relative evaluation of the cooling water chemicals, corresponding to their ratios for PEC/PNEC or PEC/EQS. A conceivable further development of the concept for the evaluation of substance mixtures exists in the introduction of an Index-Number (=sum of the PEC/PNEC-ratios of the individual components). The advantages of this immission-based procedure are that substance-inherent properties of the chemicals can be used and the local conditions taken into account. However, the focus is on the aquatic ecotoxicity. Indirect negative effects on humans are not considered, insofar as there are no official EQS derived from the chemical evaluations of the EU available. A further disadvantage is that the PEC depends on the flow of the receiving river, so that there is a tendency to view dilution as solving the problem ("dilution as solution"), while the real goal of limiting the load recedes from view. In addition, the additional burden on the receiving river through other emissions upstream can only be taken into account with great effort. While the elimination in sewage plants is fully considered, also for adsorptively bound chemicals, other exposure paths, e.g., through an agricultural utilization of the activated-sludge ase not considered. The "benchmarking"-concept thus remains only a promising approach, which still should be further developed. 7.4.3 Plant specific evaluation of cooling water chemicals The benchmarking-concept was further developed by Baltus et al., among others, especially to provide an instrument with which the use and selection of cooling water biocides could be optimized (Baltus and Berbee 1996, van Donk and Jenner 1996, Baltus et al. 1999). The basis for this is an extension of the benchmarking-concept to actual cooling plants. While in the benchmark-process the inital pre-selection of the biocide is made on the basis of relative "rankings", in the optimization phase all available techniques for process operation, dosing, treatment and control are considered. Thereafter the local effects on the environment are estimated and

96 measures for their reduction, for example through treatment of the added water, dosing, closing the outflow or monitoring, are established. 7.4.4 TEGEWA-concept for indirect dischargers The TEGEWA, together with the Gesamtverband der Deutschen Textilveredelungsindustrie developed an evaluation concept for the classification of textile treatment aids. Based on their relevance for surface waters and in 1997 in this was established as a self-commitment to be followed voluntarily (Noll and Reetz 1998). The basis is a classification of textile additives in 3 classes from slightly to strongly relevant for emitted water depending on their composition of bio-accumulable substances (classification in R 53), their toxicity for surface waters (LC 50 Daphnia), their ready bio-degradability (OECD 301 tests) and their eliminability in sewage plants (Zahn-Wellens test). Certain problem substances (cancerogenic, genotoxic and reproduction toxic substances) a priori lead to classification in Class 3, while on the other hand textile chemicals, which are hazardous to water organisms (R 52) and are not readily degradable (R 53) are still assigned to Class 1 if substantial elimination in sewage plants occurs. In its structure the TEGEWA-concept resembles a classification of water-endangering substances, but it takes into account the elimination in municipal sewage plants. 7.4.5 Outlook From the viewpoint of the researcher a combination of an emission-based approach based on the classification in water-endangering classes or the TEGEWA-concept with the immission-based "benchmarking"-approach would be a possible instrument for making a pre-selection of suitable cooling water chemicals. The assignment to water-endangering classes also includes health aspects, along with aquatic ecotoxicity. Moreover, any lacking data are compensated by default values following the precautionary principle. However, in order to place priority attention on the emitted loads, both the used amounts as well as the elimination in the cooling system and (for indirect dischargers) in the municipal sewage plant must be more strongly considered. Thus the TEGEWA-concept tends more to include the elimination in the wastewater treatment facility. The total load should, however, also be considered, so that, for example, effective biocides with higher aquatic toxicity but a lower use concentration are not a priori rated as worse than less effective biocides

97 with higher loads. The total load is included in the evaluation in the "benchmarking"concept, but is also relativized through the dilution in the final settling pond. However, the goal should be the minimizing principle, in introducing hazardous substances while at the same time maintaining the goal of environmental quality. In order to direct the idea of "benchmarking" in this direction, a relative ranking of cooling water chemicals in the direction of toxicity loads (=effect concentrations x load) should be possible. In general, the emission-based wastewater-evaluation is regarded as the stricter, precautionary principle-bound approach, while the immission-based approach first of all requires the definition of quality criteria. Of course, there have been attempts to link the two strategies. Also in the EU-Water Framework guideline the "combined approach" has elements of emission- (emission controls for prioritary substances) and immission-based criteria (environmental quality standards), (cf., also Reemtsma and Klinkow 2001). More advanced concepts for the evaluation of wastewater inputs on the basis of ecotoxicity, bioaccumulation, degradability and genotoxicity have been developed by Reemtsma and Klinkow, among others, and are currently being discussed at the OSPAR (Reemtsma and Klinkow 2001, Commission 2000). In order to undertake a qualified evaluation of cooling water chemicals their identity and properties must be known. In the Annex some essential data regarding

ecotoxicity, genotoxicity and biological degradability, which were obtained in extensive substance-based searches, are listed together with the R-Phrases and the physico-chemical properties. It should be emphasized that this data set is only sufficient to permit an initial estimation of the environmental relevance of cooling water chemicals to be made and is not sufficient for a complete evaluation in accordance with chemical laws. In any case, the optimization of the use of cooling water chemicals must be made with a consideration of the local conditions and the available techniques for process management, dosing, treatment and monitoring. The various evaluation strategies described above can only reflect general basic principles. In order to achieve the needed linkage of the emission- and immission-based approaches, these strategies must be further developed.

98

8 Sources
78/659/EWG. 1978. Guideline des Rates 78/659/EWG vom 18. Juli 1978 ber die Qualitt von Swasser, das schutz- oder verbesserungsbedrftig ist, um das Leben von Fischen zu erhalten. Abou-Ghazala, A. and K. H. Schoenbach. 2000. Biofouling prevention with pulsed electric fields. IEEE Transactions on Plasma Science 28:115-121. Allonier, A.-S., M. Khalanski, V. Camel and A. Bermond. 1999. Characterization of Chlorination Byproducts in Cooling Effluents of Coastal Nuclear Power Stations. Marine Pollution Bulletin 38:1232-1241. Andres, O., F. J. Angenend, G. Greul, A. Schiffers, K. Schluck, W. Sielhorst and H. J. Tiefenthal. 1980. Der Einsatz von Khlwasserbehandlungsmitteln in Kraftwerken und ihr Einflu auf die Umwelt. RWE-ERFA-TC Arbeitskreis conditioning. Anonymous. Entwurf einer Verordnung ber Anforderungen an das Einleiten von Abwater in Gewsser und der Anpassung der Anlage des Abwaterabgabengesetzes vom 22.10.2001, BMU WA I3-21110-1/5 Anonymous. 1991. Betz Handbook of Industrial Water Conditioning, 9th ed. Betz Laboratories, Inc., Trevose, USA. Anonymous. 1994. Biocides in Cooling Water Systems. Consultants in Environmental Sciences Limited by order of the U.K. Department of the Environment, Kent, U.K. Anonymous. 2000. Reference Document on the Application of Best Available Techniques to Industrial Cooling Systems. European Commission Joint Research Centre, European IPPC Bureau. Anonymous. 2001a. Code of practice for the control of legionella bacteria in cooling towers. Institute of Environmental Epidemiology, Ministry of the Environment, WHO collaborating centre for environmental epidemiology, Singapore. Anonymous. 2001b. "Sicherheitskonzept fr Khlwasserstrme in der Chemischen Industrie". Verband der Chemischen Industrie e. V. Anonymous. 2001c. Synthesebericht Antifoulings und Khlwassersysteme (Entwurf). Internationale Kommission zum Schutz des Rheins. Baltus, C. A. M. and R. P. M. Berbee. 1996. Het gebruik van biociden in recirculatiekoelsystemen. 96.036, Rijksinstitute voor Integral Zoetwaterbeheer en Afvalwaterbehandling RIZA, Lelystad. Baltus, C. A. M., L. C. M. Kerkum and P. G. M. Kienhuis. 1999. Acute toxiciteit van koelwaterlozingen uit recirculatiekoelsystemen. RIZA rapport 99.025, Rijksinstitut voor Integraal Zoetwaterbeheer en Afvalwaterbehandeling, Lelystad. Brnert, W. and U. Hagendorf. 1992. Biologische Testverfahren zur Feststellung gefhrlichen Abwassers im Sinne 7a WHG (Zwischenbericht fr den Zeitraum vom 01.01.1089 - 30.06.1991). F+E 102 05 147 im Auftrag des Umweltbundesamtes, Institut fr Wasser-, Boden- und Lufthygiene. Beratergremium fr Umweltrelevante Altstoffe (BUA) 1996. 2-Phosphono-1,2,4butantricarbonsure (PBTC). in BUA-Stoffberichte 180-183. S. Hirzel Verlag, Stuttgart.

99 Statistisches Bundesamt, S. 1998. Wasserversorgung and Abwaterbeseitigung im Bergbau und Verarbeitenden Gewerbe und bei Wrmekraftwerken fr die ffentliche Versorgung 1995, Fachserie 19, Reihe 2.2, Wiesbaden, Mai 1998 Cloete, T. E. and V. S. Brzel. 1992. Practical aspects of biofouling control in industrial water systems. International Biodeterioration & Bio-degradation 29:299-341. Commission, E. 1996. Technical Guidance Documents in Support of the Commission Directive 93/67/EEC on Risk Assessment for New Notified Substances and the Commission Regulation (EC) 1488/94 on Risk Assessment for Existing Substances, Part II, Chapter 3: Environmental Risk Assessment. in Office for Official Publications of the European Communities, Luxembourg. Commission, O. 2000. Ospar Background Document concerning the elaboration of programmes and measures relation to Whole Effluent Assessment. Ospar Commission, The Executive Secretary, London. Cornell, J. S., D. A. Pillard and M. T. Hernandez. 2000. Comparative measures of the toxicity of component chemicals in aircraft deicing fluid. Environmental Toxicology and Chemistry 19:1465-1472. Daamen, E. J., J. W. Wouters and J. T. G. Savelkoul. 2000. Side stream biofiltration for improved biofouling control in cooling water systems. Wat. Sci. Tech. 41:445-451. Diehl, K. and U. Hagendorf. 1998. Datensammlung Bioteste. UBA-Texte 9/98, Umweltbundesamt, Berlin. Diehl, M. A. and J. S. Chapman. 1999. Association of the biocide 5-chloro-2-methylisothiazol-3-one with Pseudomonas aeruginosa and Pseudomonas fluorescens. International Biodeterioration & Bio-degradation 44:191-199. DIN 38402. 1991. Deutsche Einheitsverfahren zur Wasser-, Abwasser- und Schlammuntersuchung. Allgemeine Angaben (Gruppe A) Probenahme von Khlwasser fr den industriellen Gebrauch (A22). Elfering, R. and T. Schmitz. 1995. Gesteuerte Khlwasserkonditionierung. Chemie Technik 24:48-49. Fichte, W., H.-J. Krabbe, H.-G. Seipp, J. Wasel-Nielen and R. Truppat. 2000. VGBKhlwasserrichtlinie "Wasserbehandlung und Werkstoffeinsatz in Khlsystemen. VGB-R 455 P, VGB Technische Vereinigung der Grosskraftwerkbetreiber e.V., Essen. Fielden, T. and M. Iddon. 1997. Use Category Document - Water Treatment Chemicals. Environmental Agency-Chemicals Assessment Unit, Ecotoxicology and Hazardous Substance National Centre, Wallingford, Oxon. Gartiser, S. and R. Scharmann. 1993. Biologische Testverfahren im Bereich des Abschlmmwassers von Khlturmkreislufen. wasser abwasser praxis:402-407. Gartiser, S. and E. Urich. 2001. Protokoll des UBA-Workshops zu F+E FKZ 2002433 "Einsatz umweltfreundlicher Chemikalien in der Khlwasserkonditionierung" am 18.05.01. Hydrotox GmbH, Freiburg. Gellert, G. and A. Stommel. 1995. Untersuchung ber die Hemmwirkung von Khlwasser auf die Lichtemission und auf das Wachstum von Leuchtbakterien. Korrespondenz Abwasser 42:1324-1333.

100 German, R., P. Scheidel, H. Steinlein and E. Weindel. 1996. Abklingverhalten von Mikrobiziden im Khlwasser - Entwicklung einer Simulationsmethode. Chemie Technik 25:52-55. Gledhill, W. E. and T. C. J. Feijtel. 1992. Environmental properties and safety assessment of organic phosphonates used for detergent and water treatment applications. Pages 261-285 in O. Hutzinger, editor. The handbook of environmental chemistry. Springer-Verlag, Berlin. Greiner, G., P. Thomas, H. Kittel and R. Truppat. 1987. Merkblatt "Abwasser und feste Rckstnde aus Wasseraufbereitungsanlagen und wasser-/dampfberhrten Systemen in Wrmekraftwerken - Anfall und Verbleib", Verlag technisch-wissenschaftlicher Schriften edition. VGB Technische Vereinigung der Frosskraftwerkbetreiber e.V., Essen. Griebe, T. and H.-C. Flemming. 1996. Vermeidung von Bioziden in Wasseraufbereitungs-Systemen durch Nhrstoffentnahme. Vom Wasser:217-230. Groshart, C. P. and F. Balk. 2000. Biocides. Association of River Waterworks RIWA, Amsterdam. Guiney, P. D., J. E. McLaughlin und J. Hamilton. 1996. Dispersion Polymers. pages 147-165 in J. D. Hamilton und R. Sutcliffe, editors. Ecological assessment of polymers. Van Nostrand Reinhold, New York. Gunten, U. v. and J. Hoign. 1994. Bromate Formation during Ozonation of BromideContaining Waters: Interaction of Ozone and Hydroxyl Radical Reactions. Environ. Sci. Technol. 28:1234-1242. Hsserbarth 1987: Schreiben des Inst. fr Wasser-, Boden- und Lufthygiene des Bundesgesundheitsamtes an Dr. Kleinstck, Bayer AG vom 23.07.1987 Held, H. D. and H. G. Schnell. 2000. Khlwasser. Vulkan-Verlag, Essen. Hild, W. 2000. Keine Chance fr Biofouling - Kontrolle der mikrobiologischen Aktivitt in Khlwassersystemen. Verfahrenstechnik 34:36-39. Holzwarth, G., R. G. Balmer and L. Soni. 1984a. The Fate of Chlorine and Chloramines in Cooling Towers. Water Res. 18:1421-1427. Holzwarth, G., R. G. Balmer and L. Soni. 1984b. The Fate of Chlorine in Recirculating Cooling Towers. Water Res. 18:1429-1435. Horstmann, B. and A. Grohmann. 1986. Der Einflu von Phosphonaten auf die Phosphatelimination. Z. Wasser- Abwasser-Forsch. 19:236-240. Howland, E. B. and D. H. Pope. 1983. Distribution and Seasonality of Legionella pneumophila in Cooling Towers. Current Microbiology 9:319-324. Jop, K. M., P. D. Guiney, K. P. Christensen and E. M. Silberhorn. 1997. Environmental fate assessment of two synthetic polycarboxylate polymeres. Ecotoxicology and environmental safety 37:229-237. Kaiser, T., W. Schwarz, M. Frost and W. Pestemer. 1998. Evaluierung des Gefhrdungspotentials bisher wenig beachteter Stoffeintrge in Gewsser. F+E-Nr. 296 72 036, Umweltbundesamt, Berlin. Klautke, S. B. 2001. Personal communications from 04.04.01. BetzDearborn, Willich.

101 Knepper, T., J. Driemler, A. Maes, J. Mller, and D. Sodorf. 2002. Eintrge synthetischer Komplexbildner in die Gewsser. Umweltbundesamt F+E-Bericht 299 24 284 UBA-Texte 03/02, Umweltbundesamt, Berlin. Knichel, H. 2001. Personel communication from 30.05.01 about cooling water supply in power plants. Statistisches Bundesamt Zweigstelle Bonn Kleinstck, D. 23. May 2001. Personal communication Dr. Kleinstck (Bayer-AG) und technische Berichte SP-IOW-W.0033/99/3 und SP-IOW-W.0034/99/1. Kleinstck, D. 07.03.2002. Personal communication Dr. Kleinstck (Bayer-AG) Klinger, J., F. Sacher, H.-J. Brauch and D. Maier. 1998. Verhalten organischer Phosphonsaeuren bei der Trinkwasseraufbereitung. Vom Wasser 91:15-27. Klinger, J., F. Sacher, H.-J. Brauch und D. Maier. 1999. Determination of organic phosphonates in aqueous samples using liquid chromatography/particle-beam mass spectrometry. Acta hydrochim. hydrobiol. 25:79-86. Kohler, R., editor. 2000. Merkblatt ATV-DVWK-M706, Teil 1 "Abwasser, das bei der Wasseraufbereitung anfllt". GFA-Gesellschaft zur Frderung der Abwassertechnik e.V., Hennef. Kommission zur Bewertung wassergefhrdender Stoffe, K. 1999. Einstufung wassergefhrdender Stoffe auf der Basis der Verwaltungsvorschrift wassergefhrdender Stoffe (VwVwS) vom 17.05.1999 (LTwS-Nr. 10). Umweltbundesamt, Berlin. Kool, H. J. and J. Hrubec. 1986. The influence of an ozone, chlorine and chlorine dioxide treatment on mutagenic avtivity in (drinking) water. Ozone Science and Engineering 8:217-234. Koppensteiner, D. G. 1973. Biologische Probleme in Khlwasserkreislufen. wasser luft und betrieb 17:424-427. Kroiss, H. and M. Zessner. No year. Treatment and sludge disposal-what are the challenges? Proceedings of the Euro-Case workshop "waste water sludge as a challenge", Austrian Academy of Sciences and the European Councli of Applied Sciences and Engenieering - Vienna 25 June 2001, Institute for Water Quality and Waste Management, Vienna University of Technology. (www.eurocase.org/Activities/wastewater_sludge_010425/Kroiss.pdf). Kunz, P. M. 1993. Wasser aus Khlwasserkreislufen. awt abwassertechnik:51-53. Kusnetsov, J. M., P. J. Martikainen, H. R. Jousimies-Somer, M.-L. Visnen, A. I. Tulkki, H. E. Ahonen and A. I. Nevalainen. 1993. Physical, Chemical and Microbiological Water Characteristics Associated with the Occurrence of Legionella in Cooling Tower Systems. Wat. Res. 27:85-90. Kusnetsov, J. M., A. I. Tulkki, H. E. Ahonen and P. J. Martikainen. 1997. Efficacy of three prevention strategies against legionella in cooling water systems. Journal of Applied Microbiology:763-768. Landesarbeitsgemeinschaft Wasser, H. 1991. Grundlagen fr die Beurteilung von Khlwassereinleitungen in Gewsser, Berlin. MacDonald, R., M. Santa and V. S. Brzel. 2000. The response of a bacterial biofilm community in a simulated industrial cooling water system to treatment with an anionic dispersant. Journal of Applied Microbiology:225-235.

102 Marczewski, G., P. Scheidel and M. Walker. 1995. Schnelle Bestimmung vor Ort. Kontrolle der mikrobiologischen Aktivitten in Khlwassersystemen. CAV 3:28-30. Mattila-Sandholm, T. and G. Wirtanen. 1992. Biofilm Formation in the Industry: A Review. Food Reviews International 8:573-603. Meerkerk, M. A. and L. M. Puijker. 1997. Onderzoek naar de herkomst van AMPA in het oppervlaktewater. RIWA Samenwerkende Rijn- en Maaswaterleidingbedrijven. Metzner, G. 1990. Verhalten von organischen Komplexbildnern auf Phosphonsurebasis in Klranlagen. in Umweltvertrglichkeit von Wasch- und Reinigungsmitteln. Mnchner Beitrge zur Abwasser-, Fischerei- und Flubiologie: 323-336, R. Oldenbourg Verlag, Mnchen Naumann, G. and D. Schimke. 1991. Laboruntersuchungen zur Korrosion von unlegiertem Stahl in Khlwssern. Werkstoffe und Korrosion 42:8-11. Noll, L. and H. Reetz. 1998. Gewsserkologisch orientierte Klassifizierung von Textilhilfsmitteln. Melliand:633-635. Nowack, B. 1998. Water Res. 32:1271-1279. Nowack, B. 2001. Determination of phosphonic acid breakdown products by highperformance liquid chromatography after derivatization. Journal of Chromatography A 942:185-190. Nowack, B. and U. Baumann. 1998. Biologischer Abbau der Photolyseprodukte von Fe(III)EDTA. Acta hydrochim. hydrobiol. 26:104-108. Nowack, B. and A. T. Stone. 1999. The influence of metal ions on the adsorption of phosphonates on to Goethite. Environ. Sci. Technol. 33:3627-3633. Nowack, B. and A. T. Stone. 2000. Degradation of nitrilo(methylenephosphonic acid) and related (amino)phosphonate chelating agents in the presence of manganese and molecular oxygen. Environ. Sci. Technol. 34:4759-4765. Opgenorth, H.-J. 1992. Polymeric Materials Polycarboxylates. Pages 338-350 in O. Hutzinger, editor. The handbook of environmental chemistry. Springer-Verlag, Berlin. Paikert, P. 1986. Verschmutzung in Kondensatoren und Khltrmen. Chem.-Ing.Tech., VCH Verlagsgesellschaft mbH, Weinheim 58:709-715. Parker, G. 1986. Molybdenum. Pages 217-235 in O. Hutzinger, editor. The handbook of environmental chemistry. Springer-Verlag, Berlin. Piegsa, W. 2001a. Persnliche Mitteilung zur Khlwasserkonditionierung in der Durchflusskhlung. Staatliches Umweltamt Herten, NRW. Piegsa, W. L. 2001b. Hinweise und Erluterungen zum Anhang 31 - Wasseraufbereitung, Khlsysteme, Dampferzeugung - der Verordnung ber Anforderungen an das Einleiten von Abwasser in Gewsser. Bund-/Lnder-Gesprchskreis 21/41 "Anhang 31 AbwV", Herten. Reemtsma, T. and N. Klinkow. 2001. Untersuchungstrategie fr gefhrliche Stoffe in Abwassereinleitungen der Industrie. F+E-Vorhaben 299 22 297, Berlin. Rice, R. G. and J. F. Wilkes. 1992. Fundamental aspects of ozone chemistry in recirculating cooling water systems -- data evaluation needs. Ozone Science and Engineering 14:329-365.

103 Roth, H., K. Lucas, W. Solfrian and F. Rebstock. 1996. Die Nutzung der industriellen Abwrme zur Fernwasserversorgung - Berichtstext. F+E 104 07 312, Umweltbundesamt, Berlin. Schaaf, M. Der LUMIStox Leuchtbakterientest zur Abschtzung der Biozidwirksamkeit in Khlwasser. Dr. Lange Anwendungsbericht Bio Nr. 105, Dr. Bruno Lange GmbH, Dsseldorf. Scheidel, R. G., H. Steinlein and W. Weindel. 1996. Abklingverhalten von Mikrobiziden im Khlwasser. Chemie Technik 25:402-407. Schmittecker, B. M., K. P. Henke and R. Bergmann. 1999. Khlwasserbehandlung mit Ozon. VGB Kraftwerkstechnik 79:82-87. Schmittecker, B. M., K.-P. Henke und R. Bergmann. 1997. Khlwasserbehandlung mit Ozon - Betriebserfahrungen aus dem Grokraftwerk-. VGB Kraftwerkstechnik 77:976. Schberl, P. and L. Huber. 1988. kologisch relevante Daten von nichttensidischen Inhaltsstoffen in Wasch- und Reinigungsmitteln. Tenside Surfactants Detergents 25:99-107. Schoester, D. 2001. Personal communication from Dr. Schiester (Thor GmbH) from 27.06.01 und Prfbericht CHE 1224-1154-042 vom 10.2.96. Schweitzer, P.-H. 1995. Behandlung von Khlwasserkreislufen mit Ozon. Landesanstalt fr Umweltschutz Baden-Wrttemberg Referat 34-Arbeitsschutz, Karlsruhe. Sommer, S. 1988. Einsatz von Serienkhltrmen. Sonderdruck aus "Technische Mitteilungen" 81:3-9. Srikanth, S. and S. G. Berk. 1993. Stimulatory Effect of Cooling Tower Biocides on Amoebae. Applied and Environmental Microbiology 59:3245-3249. States, S. J., L. F. Conley, S. G. Towner, R. S. Wolford, T. E. Stephenson, A. M. McNamara, R. M. Wdowsky and R. B. Yee. 1987. An Alkaline Approach to Treating Cooling Towers for Control of Legionella pneumophila. Applied and Environmental Microbiology 53:1775-1779. Steber, J. and W. Hater. 1997. kologische Beurteilung von Khlwasserbehandlungsmitteln. VGB Kraftwerkstechnik 77:976. Todutza, K. and H. Steinlein. 1990. Minderung des Einsatzes von Konditionierungsmitteln in Khlsystemen im Hinblick auf die Gewsserentlastung. Schlubericht zum FE-Vorhaben Nr. 102 06 502, Bayer. Landesamt fr Wasserwirtschaft, Mnchen. U. N. E. P. 1996. Screening Information data set for high production volume chemicals: PBTC. IRPTC Vol. 3, Part 1, Nr. 5, International Register of Potentially Toxic Chemicals, Geneva. U. N. E. P. ohne Jahr. Screening Information data set for high production volume chemicals: ATMP. IRPTC Vol. 1, Part 2, Nr. 4, International Register of Potentially toxic chemicals, Geneva. van Donk, M. and H. A. Jenner. 1996. Optimization of biofouling control in industrial cooling water systems with respect to the environment. 64387-KES/WBR 96-3113, KEMA Environmental Services.

104 VDI 3784. 1986. Ausbreitung von Emissionen aus Naturzug-Nasskhltrmen. Beurteilung von Khlturmauswirkungen. in Verein deutscher Ingenieure, VDI Handbuch Reinhaltung der Luft, Band 1; VDI Handbuch Energietechnik. Viera, M. R., P. S. Guiamet, M. F. L. de Mele and H. A. Videla. 1999. Use of dissolved ozone for controlling planctonic and sessile bacteria in industrial cooling systems. International Biodeterioration & Bio-degradation 44:201-207. Wasel-Nielen, J. and M. Baresel. 1997. Khlwasserbehandlung mit Ozon in der Hoechst AG im Vergleich zu anderen mikrobiziden Verfahren. VGB Kraftwerkstechnik 77:130-134. Werner, H.-P. and M. Pietsch. 1991. Bewertung des Infektionsrisikos durch Legionellen in Khlkreislufen von Kraftwerken. VGB Kraftwerkstechnik 71:785-787. Wunderlich, M. 1978a. Studie ber den Einflu von Khlwasserbehandlungsmitteln auf die Gewsser. UFOPLAN Nr. 102 04 033 im Auftrag des Umweltbundesamtes, Bundesanstalt fr Gewsserkunde, Koblenz. Wunderlich, M. 1978b. Zusammenfassender berblick zur Chemie und Biologie der Kreislaufkhlung. Deutsche Gewsserkundliche Mitteilungen 22:65-73. Wunderlich, M. 2002. Personal communication from 20.03.02, Bundesanstalt fr Gewsserkunde.

105

9 Ackknowledgements

The successful completion of the F+E-Project was only possible through the cooperation and assistance of numerous persons, whom we would like to thank here most sincerely: Mr. Mehlhorn, Fachbereichsleiter Abteilung II 3.2 "Stoffhaushalt Gewsser", Environmentbundesamt and Leiter des IKSR-Expertenkreises "Antifouling and Khlwaterkreislufe" for outstanding support, advice and coordination of the project, Mr. Seelisch, FG II 3.1 "bergreifende Angelegenheiten Gewssergte and Waterwirtschaft", Environmentbundesamt for research in internal data banks, The otehr employees of Hydrotox GmbH for conscientious performance of the tests, especially Mrs. Stiene (Daphnia-, algae-tests), Mrs. Welsch (umu test, Ames test) as well as Dr. Schnurstein (coordination in the area of mutagenicity), Mrs. Knie for the secretarial work and for correcting the manuscript, Mr. Piegsa, Staatliches Environmentamt Herten and Obmann des Bund-/LnderGesprchskreises "Khlwater", The representatives of the agencies of the Lnder and Bezirke for their support with the questionnaire on the use of cooling water chemicals in the operating facilities, Dr. Ungeheuer, TEGEWA, for the organisation of the Workshop, The employees of the facilities considered for providing information and samples and for filling out the questionnaire, The representatives of the companies manufacturing conditioners for providing product information, safety data sheets and additional, in part unpublished information, especially Dr. Klautke, BetzDearborn GmbH; Dr. Hater, Henkel Surface Technologies GmbH; Dr. Kleinstck, Bayer AG; Dr. Weindel, BK-Giulini Chemie GmbH & Co OHG; Dr. Olkis, Fa. ONDEO Nalco Deutschland GmbH; Dr. Heinl, Drew Ameroid Deutschland GmbH; and Dr. Schoester, Thor GmbH.