Brominated Flame Retardants

3. Turnover with Waste Products

3.1 Recycling of Brominated Flame Retardants
3.1.1 Waste of Electric and Electronic Equipment
3.1.2 Other Recycling Processes
3.1.3 Export of Scrap
3.1.4 Limitations on Recycling of Plastics
3.2 Disposal with Solid Waste
3.2.1 Sources of Brominated Flame Retardants to Solid Waste
3.2.2 Fate of Brominated Flame Retardants by Incineration
3.2.3 Landfilling Activities
3.3 Turnover with Chemical Waste
3.4 Turnover with Waste Water
3.5 Summary

3.1 Recycling of Brominated Flame Retardants

When products containing brominated flame retardants are discarded they will either be directly disposed of with solid waste or they will enter the recycling industry for reprocessing of valuable components.

The post consumer flow of plastics from TV sets is illustrated in figure 3.1. For TV sets and other electronics the first step when the products enter the recycling industry will usually be an initial manual dismantling.

Figure 3.1
Schematic illustration of post consumer flow of plastics from TV sets

3.1 Recycling of Brominated Flame Retardants

Processing of discarded products containing flame retarded plastics has traditionally been driven by the interest of recycling the metals.

Products containing brominated flame retardants can follow three waste fractions:

3.1.1 Waste of Electric and Electronic Equipment

Until now only a minor fraction of waste of electric and electronic equipment (EEE) has been collected and processed, but according to the new Statutory Order on managing of waste of electrical and electronic products of Dec. 1988 /49 / all EEE waste is to be processed. Plastic parts containing brominated flame retardants are to be removed from the discarded products. The flame retarded plastic is to be landfilled, incinerated or recycled for applications where flame retardancy is required.

Projections of total amount of EEE waste

Disposal of brominated flame retardants with EEE waste in Denmark have previously been covered by several analyses - all dealing with EEE waste in more general terms. In three analyses, discussed in the following, the estimate of the amount of EEE waste in Denmark has been derived from projections made by Hansen et al., 1993 /50 /. The projections are based on historic consumption figures and expected average lifetimes of the different fractions of electronic and electric products.

Brominated flame retardants in EEE waste

There is a considerable uncertainty on the total amount of flame retardants in EEE waste, especially in thermoplastics. In two studies of EEE waste, both from 1995, the total content of brominated flame retardants in plastics (exc. printed circuit boards) in EEE waste in Denmark was estimated at 1,150 and 157 tonnes, respectively.

In the following the basis for the estimates of the total amount of brominated flame retardants with EEE waste will be discussed.

In the project: 'Environmental consequences of incineration and landfilling of waste from electr(on)ic equipment' by Taberman et al. from 1995 /52 / the total amount of flame retarded thermoplastic in the EEE waste was estimated at 6,400 tonnes under the assumption that plastics accounted for 21% of the waste. Of the plastics, 26% was assumed to be flame retarded corresponding to 5.3% of the total EEE waste. This assumption is based on two German studies of EEE waste. It was moreover in the analysis assumed that the flame retarded thermoplastic on average contained 18 wt% brominated flame retardant, almost all DeBDE. Considering these assumptions the total content of PBDEs in the projected 120,000 tonnes EEE waste disposed of in Denmark in 1995 was estimated at 1,150 tonnes.

To this must be added TBBPA and PeBDE in printed circuit boards. The total amount of printed circuit boards was estimated at 3,700 tonnes with an estimated total bromine content of 75 tonnes Br. Assuming that the brominated flame retardants contain 59% bromine (TBBPA) the 75 tonnes bromine can be translated into 127 tonnes brominated flame retardants, almost all TBBPA. The total BFR content was thus estimated at about 1,280 tonnes.

In two other studies the content of brominated flame retardants was estimated by aggregating the estimates for specific groups, in which brominated flame retardants are known to be used.

The total amount of brominated flame retardants in EEE waste in Sweden has in a study of Hedemalm et al. (1995) been estimated at 334 tonnes TBBPA and 643 tonnes PBDEs. The Swedish figures were obtained by multiplying the Danish EEE waste projections of Hansen et al. figures by a factor of 1.83. Translated into Danish figures by a division by 1.83 the amount corresponds to 182 tonnes TBBPA and 351 tonnes PBDEs.

In a Danish study of Ramböll et al. from 1995 on disposal of EEE waste, the total content of flame retarded plastics (exc. printed circuit boards) was estimated at 872 tonnes /52 /. The total content of flame retarded plastic was estimated by summing up the contributions from the different product groups. The 872 tonnes flame retarded plastic was translated into some 157 tonnes brominated flame retardant under the assumption that the flame retarded plastic contains 18% brominated flame retardants. The main sources were personal and main frame computers (52-104 tonnes), TV sets (19 tonnes), coffee and tea machines (4.3 tonnes) and photocopying machines (5-10 tonnes). In addition it was estimated that the scrap contained 3,100 tonnes printed circuit board assemblies with an average content of 32 g TBBPA/kg and 1.6 g PeBDE/kg (derived from /51/) - which translate into 99 tonnes TBBPA and 5 tonnes PeBDE. The assessment does not seem not to include small flame retarded components like switches and plugs - present in a significant share of all EE equipment. The total content of brominated flame retardants will consequently be underestimated.

The first mentioned study seems to overestimate the actual amount of flame retardant in the waste whereas the two other studies may underestimate as they include only a part of the flame retarded thermoplastics.

A very accurate estimate of brominated flame retardants in EEE waste is not possible, as the consumption patterns of brominated flame retardants have changed over the years.

According to a study conducted for the Association of Plastics Manufacturers in Europe (APME) in 1992 (ref. in /53 /) 1.5 million tonnes plastics - of these 17% flame retarded - were used in EEE products manufactured in W. Europe. Of the flame retarded plastics 78% by weight was treated with brominated flame retardants. BFR-containing plastics consequently made up about 13% of the total plastic. The average content of the flame retardants was 18%. This includes brominated flame retardants in printed circuit boards and electronic components.

In a material specific account of plastics used in the manufacturing of EEE products in Germany 1991 by the German Electrotechnical and Electronic Association, flame retarded plastic (exc. PVC) accounted for 18% of the total. Of this printed circuit boards and thermosets (excel. printed circuit boards) account for 25% and 17% respectively. The account may according to the authors not include all applications, but seem to be well in agreement with the figures from APME.

If it is assumed that 13% of the plastic contain brominated flame retardants and 50-60% of this is thermoplastics, BFR-containing thermoplastic will make up 6-8% of the total plastic content. With the projected 125,000 tonnes and an assumed plastic content of 21% (exc. printed circuit boards), the total consumption of flame retarded plastic would be around 1,800 tonnes, translating into 330 tonnes brominated flame retardants.

The amount of office equipment waste is as discussed in the following significantly overestimated in the projection. As office equipment represents a major part of the total amount of brominated flame retardants, an approach where the amount of brominated flame retardants is estimated from total EEE waste figures is not feasible.

Consequently the approach of Hedemalm et al. /32/, adding up the contributions of each product group, is applied.

It has not been possible to receive more specific information on the content of small flame retarded plastic parts of each subgroup. Nevertheless, to take the many small contributions into account, a rough estimate is added for each product group, based on present day consumption figures.

Uncertainties on the amount of EEE waste

Beside the uncertainty on the average content of flame retarded plastic in the EEE waste there is also some uncertainty on the estimate of the total amount of EEE waste.

Based on Hansen et al. (1993) the total EEE waste in 1997 can be estimated at 125,000 tonnes /50/; corresponding to 25 kg per capita.

Actual collected amounts Based on experience from a recycling plant covering around 4% of the country, only 1.5 kg electronic waste per capita was collected via the municipal collection systems in 1996 /54 /. Extrapolated to the whole country it corresponds to roughly 7,800 tonnes/year. The waste collected did not include large domestic appliances.

If large domestic appliances and lighting equipment are excluded from the projected 125,000 tonnes, still some 69,000 tonnes scrap of electric and electronic products are projected. This corresponds to roughly 13 kg per capita per year.

There may be several explanations of the difference between the predicted quantities and the actual collection.

Uncertainties on the amount of EEE waste

There is a significant uncertainty in the projections due to the used life-span estimates. Especially for computers and other office machines where there has been a steep increase in consumption the projections are very sensitive to life-span estimates.

In addition, the projected number of discarded computers is very high compared to the actual consumption. It is projected that 1.2 million computers should be discarded in 1997. A market analysis indicates that in total only 0.6 million computers including notebooks, workstations, and servers were sold in Denmark in 1997.

The total amount of office electronics scrap in 1997 is projected to be around 25,000 tonnes. An analysis performed by the trade organisation for offices and data estimates that the total amount of discarded IT equipment (office machines) will increase from around 4,000 tonnes in 1993 to around 6,000 tonnes in 1998 (ref. in /55 /).

Based on this information the amount of office instruments in EEE waste will roughly be estimated at only 25% of the projected.

The amount of discarded consumer electronics is projected to be around 12,000 tonnes in 1997 /50/. An analysis performed by the trade organisation for consumer electronics estimate the total amount in 1997 to be around 11,000 tonnes - concurrent with the projection (ref. in /55/). An explanation of the differences between the quantities received by the recycling plant mentioned above and the total projected amount of EEE waste could be that the recycling plant mainly receives consumer electronics, whereas other product groups are still disposed of with waste to landfills and incineration.

In total around 17,000 tonnes of the electronic waste with relatively high content of brominated flame retardants were discarded in 1997.

Amounts of brominated flame retardants in EEE waste

An estimate of the amount of brominated flame retardants in EEE waste is carried out in appendix 6 and is included in table 3.2.

The total content of brominated flame retardants in electronic appliances and electric machines are estimated using the same groups as in the section 2.2: Printed circuit boards, housing and other parts.

In a German study Riess et al. (1998) analysed 78 TV housings and 34 PC housings from a recycling company for the presence of flame retardants /56 /. PBDEs were present in 78% of the samples, PBBs in 16%, 1,2-Bis-(tribromophenoxy)ethane in 3%, and other flame retardants accounted for the remaining 3%. It is not stated in the article, but the PC housing is presumed to represent the monitor housing. The relatively high share of the housing containing PBBs reflects that ABS some years ago was the main application for PBBs /6/.

According to the experience of a Swedish recycling company 80-90% of all PC monitors, 60-70% of TV back plates, and a minor part of other consumer electronics contain BFRs /38/.

Laser printers and photocopying machines of today contain BFRs. It can roughly be assumed that all discarded printers, wordprocessors and photocopying machines contain BFRs.

The amount of other plastic parts of appliances, lighting, wiring parts, and equipment for industrial automation will be estimated from present day consumption figures.

Processed EEE waste In 1997 only 3-4,000 tonnes EEE waste was selectively processed for recycling in Denmark. TV-sets and consumer electronics accounted for the major part of the processed scrap.

The flow of flame retarded plastic through the process used by one of the Danish recycling plants is illustrated in figure 3.2. The flow diagram is estimated to be representative to Danish recycling plants.

At the recycling plant large plastic parts like monitor housing is removed. About 5% of this plastic - mainly transport packaging and cable reels - is recycled, whereas the remaining plastic is disposed of to solid waste incineration. None of the recycled plastics contain brominated flame retardants.

Small plastic parts are left on the appliances for further processing in a shredder. By the shredding process plastics end up in a residue fraction that is landfilled.

Cables are removed from the products and sent for reprocessing at a cable recycling plant. The insulation of wires from EEE waste is a mixture of many types of plastic and rubbers, and recycling is not feasible. Therefore the plastic and rubber from the cables are landfilled.

Figure 3.2
Flow of flame retarded plastic through the process used by Elektro-Miljø A/S /57 /

Recycling of plastics During recycling of plastics containing brominated flame retardants, brominated dibenzofurans and brominated dibenzo-p-dioxins may be formed. The information available on the levels of dibenzofurans and brominated dibenzo-p-dioxins in PBDE containing plastics during recycling indicates that the levels within the plastics do not increase during the recycling process /60/. Results on PBDD/PBDF formation during recycling of TBBPA containing thermoplastics showed that the total PBDD/PBDF concentration in the plastics was below 5 mg/kg even after 5 recyclings (ref. in /7/.

3.1.2 Other Recycling Processes

Shredding

The first step in the recycling of vehicles and large domestic appliances is a fragmentation in a shredder, where the metals are separated from other materials. Plastics, rubber, paper, wood, dirt, etc. end up in several fractions of shredder residues. The plastic parts mainly end up in a fraction called "fluff".

A number of experiments on incineration of shredder waste has been performed in Denmark, but at present the fluff is ordinarily landfilled /58 /.

The number of cars in Denmark is increasing at the moment, and consequently the number of scrapped cars is somewhat smaller than the number of new cars. The uncertainty on the estimate of the content of brominated flame retardants in scrapped cars and other vehicles is, however, primarily a result of the uncertainty on the estimate of the content of brominated flame retardants in cars produced 10-15 years ago.

In the absence of specific information it is roughly estimated that an amount corresponding to 80% of todays consumption with vehicles is discarded. This corresponds to 27 tonnes flame retardants.

Large domestic appliances are also usually fragmented in a shredder.

Steel production Small automotive plastic parts containing brominated flame retardants may follow scrap steel to secondary steel production. Brominated flame retardants are generally not used in fire resistant paint for steel construction. It is not possible to estimate the amount of brominated flame retardants in plastic going into steel production, but it is roughly estimated to be less than 1% of the amount used in vehicles corresponding to less than 1 tonne/year. The plastic parts will be combusted during the melt down.

3.1.3 Export of Scrap

Printed circuit boards from processed EEE waste were exported for recovery of the metals. The printed circuit boards were exported to W. European countries; among others Sweden and Germany. Processes for recovery of the metals have been described by Legarth 1996 /33/. As the first step of the recovery the plastics containing brominated flame retardants are often removed from the scrap by a pre-treatment incineration.

The treated scrap can roughly be estimated to represent 10% of the total turnover of printed circuit boards corresponding to 11-16 tonnes BFRs. Though treated scrap represents less than 10% of the total EEE scrap, it is estimated that the treated scrap represents a larger part of the electronics with a high content printed circuit boards.

Beside this, estimated 1,000-2,000 tonnes EEE waste were exported for reprocessing in other W. European countries. This waste is roughly estimated to account for another 10-20 tonnes BFRs.

Ships and aircraft are exported for recycling abroad. It is very difficult to estimate the total content of brominated flame retardants in these products, as ships and aircraft are relatively old when they are discarded. Considering todays consumption of flame retardants with these products the total amount is roughly estimated to be <15 tonnes per year.

In total it is estimated that around 20-50 tonnes of BFRs are exported with waste products.

No import of flame retardants with waste products has been identified.

3.1.4 Limitations on Recycling of Plastics

There is at present no recycling of plastics containing brominated flame retardants from discarded products.

Recycling of brominated flame retardants with thermosets (e.g. epoxy, polyurethane and unsaturated polyester) is limited by the fact that thermosets in general are not recyclable. Thermosets may be recovered by an alcoholysis where the thermosets are split into their monomers /33/, but this process is not used at the moment. More details can be found in /33/.

For high volume thermoplastic applications like TV back panels and PC-monitor casings, recycling may have a potential, but the recycling is restricted by the fact that the casings are made of a variety of base polymers (e.g. ABS, PC/ABS, HIPS) and contain different additives. The plastics may be downcycled and used for some products like park benches or noise shields, but without a thorough separation of the plastic, brominated flame retardants and other additives would be spread without any control.

For components made of flame retarded engineering plastics like PBT and polyamide, the plastic parts are moreover so small that a separation by means of the current technology is not a paying proposition.

Recycling of flame retarded plastics may be enabled by equipment that can analyse every single item for the presence of bromine and other elements. Such equipment is used by Swedish recycling plants and is planned to be introduced in Denmark in 1999.

3.2 Disposal with Solid Waste

Practically all flame retarded plastic from discarded products will ultimately be disposed of by incineration or landfilled.

The most recent Danish assessment of waste fractions by disposal method refers to 1996 (see table 3.1). Of the 84% domestic waste that was disposed of to incineration or landfilling, 10% was landfilled and the remaining 90% was incinerated. Calculated in the same way, 53% of the bulk waste, that was not recycled, was landfilled and the remaining 47% was incinerated.

These figures will be used for an estimate of disposal of brominated flame retardants.

Table 3.1
Waste fractions by disposal method in Denmark 1996 /59 /

The numbers in table 3.1 represent averages. For many waste categories the disposal method will depend on the combustibility of the waste. Combustible building waste and office waste will thus usually be disposed of to incineration, if there is available incineration capacity. The incineration capacity is indicated by the fact that 90% of the not recycled domestic waste is incinerated.

Accordingly it will be assumed that also 90% of the combustible building and industrial waste that is not recycled will be incinerated.

Office waste will be included in more of the waste fractions in table 3.1. From the waste statistics (not shown) it appears that 61% of the waste from institutions, trade and offices were disposed of to incineration or landfilling - of this 26% was landfilled and the remaining 74% was incinerated.

The estimated disposal method for the different groups of BFR containing waste is shown in table 3.2.The actual disposal method has been changing fast during the nineties, and for most of the product groups the distribution is based on rough estimates of most probable disposal methods.

Table 3.2
Estimated disposal methods for BFR containing solid waste

3.2.1 Sources of Brominated Flame Retardants to Solid Waste

Estimates on the total content of brominated flame retardants in solid waste are shown in table 3.3.

The amount of flame retardants in the individual waste groups is very uncertain and only best estimates are shown. Considering the uncertainty of the individual groups of waste the total amount of brominated flame retardants in the solid waste is estimated at 260-560 tonnes.

EEE waste For estimation of the total BFRs in printed circuit boards a calculation similar to the calculation of BFRs in new printed circuit boards in table 2 in appendix 6 was carried out (data not shown). For the estimation of the number of disposed units 1997 projections of Hansen et al. were used. To account for the overestimate of the number of office electronics, the projection for office electronics was multiplied with 0.33. The total TBBPA content of the assembled boards of electronics other than consumer electronics is according to Hedemalm et al. /32/ estimated at 490 g/m2. Boards of TV set is estimated to contain 136 g TBBPA per m2, whereas the boards of other consumer electronics are estimated to contain 90 g TBBPA per m2 and 36 g PBDEs per m2.

For estimation of the total BFRs in housing a calculation similar to the calculation of BFRs in housing of new electronics in table 1 in appendix 6 was carried out (data not shown). For the estimation of the number of disposed units 1997 projections of Hansen et al. were used. To account for the overestimate of the number of office electronics the projection for office electronics was multiplied with 0.33. The content of BFRs in housing was estimated from the information on housing of discarded electronics in section 3.1.1.

For the group 'Control and process equipment and other equipment and wiring' the estimate has been roughly based on the present consumption (section 2.2.5) assuming that the amount of discarded equipment corresponds to 30-60% of current consumption. For cables the amount of discarded products is assumed to correspond to 80-100% of the current consumption.

BFRs in discarded small and large domestic appliances and electric tool have been roughly estimated considering the current consumption of BFRs with other part of electric and electronic equipment in section 2.2.3. For lighting it is roughly assumed that BFRs in discarded products equal the current consumption with lighting.

The basis for the estimates of the other groups than EEE waste appears from the respective sections in chapter 2.

EEE waste represents about 78% of the total content of brominated flame retardants in the waste.

Other waste groups of significance are vehicles (7% of total), building materials (6%), and waste from production of printed circuit boards (7%).

Table 3.3
Total content of brominated flame retardants in solid waste (best estimates)

3.2.2 Fate of Brominated Flame Retardants by Incineration

By combining the information in table 3.2 and 5.3 it can be estimated that a total of 170-360 tonnes brominated flame retardants were disposed of to incineration in 1997.

The total bromine content of the flame retardants can be estimated at 100-250 tonnes Br.

The total amount of incinerated waste was 2.5 million tonnes. The flame retardants contributed to a bromine content of 0.004-0.01 % (40-100 mg/kg).

Other bromine sources

On a global basis the brominated flame retardants account for about 20% of the bromine consumption. Of the other applications (see section 1.3) gasoline additives, agrochemical, etc. representing at least 50% of the total consumption will not end up in solid waste. Bromine in trace amount will be present in all kind of waste.

If it is assumed that the flame retardants account for about half of the bromine in the waste the bromine content is 0.01-0.02 %.

Other studies In foreign studies total bromine levels of 0.0015% (15 mg/kg) and 0,002-0.009% have been reported (ref. in /60 /).

By comparison the average chlorine content in the municipal waste in Denmark has been estimated at 0.25% Cl /64/.

Analysis of flue gas and residues

No analysis of brominated flame retardants or elemental bromine in flue gas or residues from Danish municipal solid waste (MSW) incinerators have been identified.

Hence, it is not possible to confirm the estimated amounts of flame retardants in solid waste by total estimated on the outflow of bromine from the incineration plants.

Emission to flue gas

At the operating temperatures of MSW incinerators almost all flame retardants will be destructed. Analyses of other organic compounds, however, demonstrate that trace amounts of the compounds pass the combustion chamber /61 /.

By way of example Hoffmann 1996 /62 / has estimated by comparing the concentration of the phthalate DEHP in the flue gas and residues from the incineration with estimates on the total DEHP content of the waste that approximately 0.1% of the total content of DEHP passed the combustion chamber and ended up in flue gas and residues.

No data on the distribution of BFRs between cleaned flue gas and the residues are available.

By comparison almost 99.9% of dioxins in flue gasses are deposited on particles that are separated during flue gas cleaning (Brna and Kilgroe 1992 ref. in /64/).

If for a worst case estimate it is assumed that 1% passes the combustion chamber and 1% of this passes the flue gas cleaning and is emitted with the flue gas to the atmosphere, the total emission of BFRs from MSW incinerators can be estimated at <0,04 tonnes per year.

The lack of data on brominated flame retardants in residues from municipal incineration is a reflection of the fact that the key issue concerning incineration of brominated flame retardants is the formation of transformation products.

Formation of dioxins and furans

There is a large body of literature that shows that polybrominated dibenzofurans and dibenzo-p-dioxins can be formed from PBDEs, PBBs and TBBPA under certain combustion/pyrolysis conditions.

The formation of polybrominated dibenzofurans and dibenzo-p-dioxins (PBDDs/PBDFs) has recently been reviewed under the International Programme on Chemical Safety (IPCS) /10/ and in the still unpublished appendix A to the EU Risk assessments of PBDEs /60/. A comprehensive review of the data on the formation of PBDDs/PBDFs during waste treatment can be found in the IPCS report /10/, and only a short introduction will be given here.

Incineration PBDDs/PBDFs may be formed during incineration by two mechanisms:

Figure 3.3 illustrates how polybrominated PBDDs/PBDFs are formed from pure polybrominated diphenyl ethers by thermal reaction involving free radical mechanism.

Figure 3.3
Formation of polybrominated dibenzo-p-dioxins (PBDDs) and polybrominated dibenzofurans (PBDFs) from PBDEs (from /63 /)

Reaction (1) occurs in pure thermal conditions (D) and reaction (2) in oxidative ones (D and O2).

Most laboratory experiments have shown that the dioxin formation and destruction depend on the incineration conditions, for example temperatures, oxygen percentage and catalysts. At temperatures of about 300°C the dioxin formation is maximal, and at temperatures above 900-1000°C the dioxins will be degraded again. The dioxin formation may occur in the combustion chamber, but is more likely to occur afterwards during the cleaning and cooling of the flue gasses /64 /.

Formation from precursors may be decreased by efficient combustion in the after-burner, whereas the de novo synthesis cannot be limited by combustion technology but instead with rapid cooling of the gasses and filter systems /64/.

Only a few studies of the effect of addition of plastics containing brominated flame retardants to incinerators have been carried out. Data from MSW incinerators in the Netherlands did not show any significant relationship between dioxin formation and the bromine content of the waste. The bromine content varied between 0.1 and 0.9% of the chlorine content /ref. in 60/.

Analyses are complicated by the fact that polychlorinated dibenzodioxins and dibenzofurans (PCDDs/PCDFs) in general are formed in much higher amount in MSW incinerators than bromo or mixed bromo and chloro dibenzodioxins and dibenzofurans (PXDDs/PXDFs) /65 /. The formation of the mixed forms PXDDs/PXDFs are explained by the extensive bromine-chlorine exchange reactions observed under several tests conditions /10/.

There are some report on the consequences of an increase in bromine input during test operations in incinerators.

In a large-scale experiment at the municipal waste incinerator at Bielefeld-Herford (Germany) material containing 4.8% PeBDE was added to the normal fuel (Lahl et al 1991, ref. in /10/). The fly ash from the electrostatic precipitator was analysed for PCDDs/PCDFs, PXDDs/PXDFs and inorganic bromine. Of the mixed PXDDs/PXDFs congeners only monobromo polychlorinated congeners could be detected. The concentrations ranged from 1.5 to 10.2 mg/kg. The concentration of the purely chlorinated compounds was much higher after the addition of PeBDE than normally detected. The results indicate that the chlorinated compounds may be formed from brominated precursors.

Wanke et al. (1996) (ref in /10/) studied the influence of additional input of BFR containing extruded polystyrene foams and rigid PUR foams to MSW incinerator at a pilot plant. No increase in PCDDs/PCDFs was detected compared with "normal fuel" in the raw gasses in the PUR foam combustion. For both foam types elevated levels of PXDDs/PXDFs were shown. The concentration of PBDDs/PBDFs was low in all experiments.

It must be expected that the formation of dioxins and furans will be dependent on the chemical structure of the flame retardants. There are no experiments where the formation of dioxins and furans after addition of different brominated flame retardants and inorganic bromine is compared.

As discussed by Jensen 1997 /64/ there is still some controversy about the proper incineration conditions for effective destruction of precursors and dioxins and furans.

3.2.3 Landfilling Activities

By combining the information in table 3.2 and 5.3 the total amount of brominated flame retardants disposed of to landfills in 1997 can be estimated at 90-200 tonnes.

Brominated flame retardants may be present in leachate from landfills, but no studies on the fate of brominated flame retardants in landfills have been carried out.

Although leaching of the compounds from plastics on a short-term scale is small, the compounds will sooner or later be released from the plastic; at least at the rate the plastic is degraded. The time scale may be hundreds of years. The question is whether the compounds are degraded before they will end up in the leachate.

As some of the compounds are persistent in the environment long term diffuse emissions from landfills cannot be neglected.

3.3 Turnover with Chemical Waste

The turnover of brominated flame retardants with chemical waste is considered to be insignificant.

3.4 Turnover with Waste Water

There is no Danish analyses of brominated flame retardants in waste water or sewage sludge.

Analyses of sludge

No analyses of brominated flame retardants in waste water have been found in the literature; only analyses of sludge.

Due to the physico-chemical properties of the brominated flame retardants the major part of flame retardants in the waste water will be adsorbed to the particles and end up in the sludge.

Two pooled samples of sewage sludge from a Swedish sewage treatment plant were found to contain 13-15 mg total PentaBDE/kg dry weight (d.w.) and 9-10 mg 2,2',4,4' TetraBDE/kg d.w /66 /. The two samples were pooled from a 32-day period with little rain and a rainy period of 7 days, respectively. There were no differences between the two periods which may indicate that the sources were primarily household and industrial effluents. The study did not include analyses of higher PBDE congeners.

The levels of the PBDEs from triBDE to hexaBDE in sewage sludge from 13 waste water treatment plants in Germany have been reported. The measured levels per kg d.w. were 0.1-1 mg triBDE, 0.2-7.5 mg tetra-BDE, 0.2-7.5 pentaBDE, and 0-1.2mg hexaBDE. In total the concentration of the four PBDEs was <17.2 mg/kg d.w.

TBBPA was found in two sewage sludge samples from two Swedish waste water treatment plants receiving leachate from a landfill with TBBPA containing industrial waste. The two single measurements were 56 and 31 mg/kg d.w., respectively /67 /.

By way of comparison PCB (polychlorinated biphenyls) levels in the sludge from three Danish waste water treatment plants ranged from 4-7 mg/kg dry weight for triPCB to 130-190 for pentaPCB /68 /.

Total production of sludge in Denmark

The annual production of municipal sludge in Denmark is around 170,000 tonnes dry weight /69 /. An order of magnitude of the total turnover of brominated flame retardants with municipal sludge can be estimated using the highest values measured in Sweden.

Estimates on BFRs in sludge

If the sludge on average contained <100 mg TBBPA/kg d.w., the total content of 170,000 tonnes would be <17 kg.

If based on the Swedish and German measurements it is assumed that the total concentration of the congeners from triBDE to hexaBDE is <30 mg per kg d.w. the total content would be <5 kg. The congeners triBDE to hexaBDE are all part of the commercial PeBDE.

There are no measurements of the most used DecaBDE in municipal sludge.

Sources According to the estimates of the EU risk assessment the by far most predominant source of DeBDE release to waste water in the EU is textile washing /3/, and the amount in waste water will thus be highly dependent on the use of BFR treated textiles. Next to this source release (via air) from polymer compounding and conversion are estimated to be the most significant source.

Industrial sources There is no data on brominated flame retardants in waste water effluents from Danish industries. Discharges of brominated flame retardants to waste water from BFR production sites have been assessed by /70 /, but from the plastic processing industries, the data were not sufficient to make any calculations of the discharge of BFRs.

Significance of releases

In a study from the Swedish River Viskan, fish and sediment samples collected upstream and downstream from several industrial point sources were analysed for PBDEs and HBCD /71 /. The lowest PBDE and HBCD concentrations were found upstream the industries and the concentration generally increased as more industries were passed. Several textile industries along the river have used DeBDE, and at least one textile factory has used HBCD. The results indicate significant release of brominated flame retardants from the textile industry, but no samples of waste water effluents from the industries were analysed. In Denmark no use of brominated flame retardants in the textile industry has been identified, and the release of brominated flame retardants from the textile industry is assumed to be insignificant.

TBBPA and its demethylated derivative MeTA were analysed in sediment samples upstream and downstream from a Swedish plastic factory using TBBPA /67/. The concentration of TBBPA was upstream and downstream 34 and 270 ng/g d.w., respectively, whereas the concentration of MeTA was 24 and 1,500 ng/g d.w., respectively. Similar differences were found when the concentration of the compounds was normalised to the content of organic matter (measured as ignition loss). Only single samples were taken and the significance of the measurements was not determined. The measurements, however, indicate significant releases of TBBPA from the factory.

Model estimates

Release to waste water from processing of plastics will depend on the actual processes. In the absence of actual measurements the possible emissions will be estimated from the model developed in 'Use category document. Plastic additives' /25/.

For organic flame retardants it is in the model assumed that initial losses from plastic processing (conversion) will be to the atmosphere. Subsequent condensation could result in losses to liquid waste. On the basis of volatile loss factors for worst case conditions, the loss for closed processes, which would be most common, is estimated at 0.002%. For processing significantly in excess of 200°C and for smaller processing sites (<750 tonnes plastic per year) loss factors should be increased by 10x. A calculation based on a loss factor of 0.002-0.02% and an annual consumption of BFR for production in Denmark of 130-190 tonnes, gives a total emission to the air of 2.6-38 kg BFR per year. Only a minor part of this can be assumed to be condensed and released to waste water. For a worst case estimate it will be assumed that half of the released flame retardants ends up in waste water corresponding to <19 kg. The emission to the air will be estimated at the calculates <38 kg.

Particulate emission of flame retardants from compounding of plastics is in /25/ estimated at 0.01-0.05% depending on particle size. The particulates will ultimately end up in solid waste or waste water. The exact amount of BFR used for compounding in Denmark is not known, but under the assumption that the 29 tonnes imported as chemicals are used for compounding a total loss of 2.9-14.5 kg BFR can be estimated. The loss is assumed to be disposed of with solid waste, but a minor part could be released to waste water.

In total it is estimated that the release to waste water from conversion and compounding is <25 kg. Of this PBDEs may account for about 1% and this source is consequently insignificant compared to the discharge from washing activities (mentioned in the next).

The major part of brominated flame retardants is used in electronic and electric products from which direct release to waste water is assumed to be negligible. BFR-containing products in direct contact with water are roofing and textiles.

When textiles containing flame retardants are washed the flame retardants will gradually be lost. It has been reported that fabrics treated with DeBDE and antimony trioxide decreased significantly in oxygen index beyond 15 launderings, indicating that the flame retardants leached from the fabrics /3/.

DeBDE has been used for protective clothing that are still in use. In section 2.2.6 it is estimated that around 50 kg DeBDE or lower PBDEs are likely to be released per year to the waste water from uniforms through the washing processes. No other uses of brominated flame retardants in clothing have been identified, but it is likely that brominated flame retardants are present in protective clothing for some civil applications. Based on the present information the release of PBDEs from clothing is roughly estimated at 50-200 kg per year.

Release of DeBDE from furniture fabrics is in the EU Risk Assessment calculated from a worst case assumption of a washing rate of once per year and assuming that 45% of the flame retardant is removed after 15 washes /3/. Based on these assumptions, the release to waste water from textiles in the UK can be estimated at 360 tonnes DeBDE in the worst case. In Denmark upholstered furniture containing brominated flame retardants is generally not used in private residences, and the fabrics are for most applications not removable. Release to waste water from washing of the fabrics is thus not very likely, and compared to the release from the washing of clothing the release is estimated to be small.

Brominated flame retardants in polyolefins used for roofing may gradually be released to rainwater. No actual release data exist. In the absence of data emission factors from 'Use category document. Plastic additives' /25/ will be used. Using the worst case emission factor for outdoor uses of 10 per year and assuming that the accumulated quantity corresponds to 10 years of the present consumption, the maximal release is estimated at 280 kg per year. On the roof the compounds will be exposed to high UV radiation levels, and a photolytic degradation of the flame retardants on the surface of the roofing is likely to take place.

Rain water PBDEs are very stable compounds, and the major part of PBDEs emitted to the air is likely to be deposited on land or the sea by dry or wet deposition (see. section 1.4). Other brominated flame retardants are more likely to be degraded before depositing. There are no deposition data on brominated flame retardants available. In section 1.4 it is estimated that 0.2-1.5 tonnes brominated flame retardants were emitted from Danish sources in 1997. Only a few percentages of the rainwater falling in Denmark will end up in waste water treatment plants and the supply of PBDEs to waste water treatment plants by rain water is roughly estimated at <2% translating into 50-200 kg per year.

Based on an estimated daily intake of 0.2-0.4 mg PBDE/day (see section 2.3), the total release of PBDEs with human excretions to waste water can be estimated at 0.4-0.8 kg/year. There are no available data on the excretions of other brominated flame retardants.

There are no measurements of leachate from landfills available.

Summary

Sources of brominated flame retardants to waste water are summarised in table 3.4. Under the assumption that the sources marked with ? is of less significance, the total discharge of brominated flame retardants to waste water is estimated at 50-530 tonnes per year. This amount agrees quite well with the estimates above based on sludge analyses from Sweden and Germany.

Table 3.4
Potential sources for brominated flame retardants in waste water

Due to the physico-chemical properties of the compounds most of the flame retardants will follow the sludge, and based on studies of other substances e.g. PCBs /68/ it will be roughly estimated that <10% will follow the water phase and be discharged to the aquatic environment. Consequently the total release from waste water treatment plants is estimated at 5-53 kg per year.

There may also be a release with urban run-off discharged directly to the aquatic environments. The main source of brominated flame retardants in the run-off is estimated to be emission from products and production processes and the release via the urban run-off will consequently be insignificant compared to the direct wet and dry deposition. It will roughly be estimated that the urban run-off maximally is half of the discharge to waste water treatment plant corresponding to <80 kg BFRs.

Sludge on agricultural soil In 1996 (most recent statistics) 69% of the municipal sludge was spread on agricultural soil /72 /. Using the same percentage for 1997 estimated 31-330 kg BFRs were spread on the soil.

3.5 Summary

The present information on the turnover of brominated flame retardants with waste products is summarised in table 3.5. It should be emphasised that the emission estimates are based on theoretical considerations.

Table 3.5
Turnover of brominated flame retardants with waste products in 1997