| Front page | | Contents | | Previous | | Next |

Assessment of the consequences of a possible ban on phosphates in household detergents

2 Environmental assessment of substances for substitution of phosphate

• 2.1 Screening for potential environmental hazard
  • 2.1.1 Degradation factor
  • 2.1.2 Release from wastewater treatment plant
  • 2.1.3 Risk index for the aquatic environment
• 2.2 Substances with high environmental hazard potential
• 2.3 Derivation of PNEC
• 2.4 Derivation of PEC
  • 2.4.1 Use of EUSES
  • 2.4.2 Emission/inventory and ECO Lab model
• 2.5 Aquatic risk assessment
• 2.6 Conclusions

The environmental assessment of the substances for substitution of phosphate in laundry detergents is carried out by a stepwise process consisting of an initial screening of the alternatives identifying the substances with the highest hazard potential, determination of predicted environmental concentrations, PEC and predicted no effects concentrations, PNEC and finally determination of risk quotients, RQ for substances identified in the screening assessment. In the derivation of RQ, the PEC is estimated by use of two concepts, EUSES and ECO lab modelling and the PNEC is calculated based on ecotoxicological studies and derived assessment factors for the individual substances.

The risk assessment is limited to cover the aquatic environment as no effect to sediment or soil organisms are known for the phosphate substitutes.

2.1 Screening for potential environmental hazard

An initial screening of the possible candidates for the substitution of phosphate is conducted. The aim of the screening is to identify substances with the highest environmental hazard potential based on the calculated risk index (R). The screening involves an initial calculation of how much of the individual substances will enter and leave the sewage treatment plant. The amount entering (IN) is dependent on the dosage of the product (e.g. weight of laundry detergent per wash) and the percentage of the substance in the product. In general, the amount leaving the sewage system (OUT) will depend on the removal taking place in the sewage treatment plant e.g. biological degradation, evaporation, sorption etc. In order to ease the screening only removal from biological degradation in the wastewater treatment plant is considered here and a degradation factor describing the biodegradability is used to estimate the effluent from the wastewater treatment plant.

The risk index is calculated based on the amount leaving the sewage system and a toxicity factor, which in the screening will be derived from the Detergent Ingredients Database list (DID-list; www.ecolabel.dk). In the calculations of the PNEC values to be used in the environmental risk assessment ecotoxicological data from existing databases (US-EPA, IUCLID and QSAR calculations) as well as an environmental risk assessment previously conducted by DHI (Danish EPA, 2001) will be used in order to use data for the exact substances where possible.

2.1.1 Degradation factor

The degradation factor for each substance is set according to the current criteria listed for the “EU-Flower” and the “Nordic Swan” ecolabel (www.ecolabel.dk). According to the ecolabel criteria, the degradation factor is 0.05 if a compound is readily biodegradable, 0.5 if the compound is inherently biodegradable and 1 if a substance is persistent.

Table 2-1 Degradation factors (DF)

¹⁾ DID list, 2007

²⁾ Danish EPA, 2001

³⁾ Dissolvin, 2007

⁴⁾ Based on evaluation of chemical structure

Two CAS numbers are listed for sodium citrate, where no data was found for the monoform and therefore data for disodium and trisodium citrate are presented. The CAS numbers presented for phosphonate are taken from the Swedish report “Fosfater i tvätt- och rengöringsmedel” (Swedish Chemicals Agency, 2006) as the representatives of phosphonates for this group.

The degradation factor for GLDA is for the L isomer only as according to the information derived from Dissolvin (2007) only this form and not the D-form is biodegradable.

2.1.2 Release from wastewater treatment plant

In order to evaluate the risk index of the use of a product an estimation of the amount of the substance in the effluent of the sewage treatment plant is done. It is assumed that the dosage (X) is 70 g per wash of laundry and the dosage is 22 g per automatic dish wash (See Appendix A). The amount of a substance entering the sewage treatment plant per wash can be calculated as followed:

IN (g) = X (g) × conc. of substance in the detergent (%)

And the amount leaving the sewage treatment plant per wash can be calculated as followed:

OUT (g) = X (g) × conc. of substance in the detergent (%)*DF

Using the typical concentration (%) listed in Table 1-2, the results listed in Table 2-2 are obtained.

Table 2-2 Calculated amounts (g) entering (IN) and leaving (OUT) the sewage treatment plant

2.1.3 Risk index for the aquatic environment

In order to evaluate which substances are associated with the highest environmental hazard potential when substituting phosphate, a risk index (R) is calculated.

The risk index (R) is given as: R =  OUT/TF

The TF is taken from the DID-list in this pre screening as mentioned previously (Section 2.1).

Substances with the highest R-value are considered to have the highest hazard potential for adverse effects when entering the aquatic environment through a sewage treatment plant.

Table 2-3 Toxicity Factor (TF) and risk index (R) for phosphate substitutes

¹⁾DID list, 2007

* TF for citrate

** TF for carbonates

*** TF for silicates

**** TF for trisodium methylglycin diacetate

No data were available for GLDA and therefore no R is calculated.

2.2 Substances with high environmental hazard potential

Table 2-4 lists the substitutes for phosphate after their potential to cause adverse effects in the aquatic environment as indicated by use of the values in the DID-list. According to the results of the screening in this study the substances with the highest risk index (R) are sodium silicate, sodium bicarbonate, sodium carbonate and zeolite for the substitution of phosphate in laundry detergents. For automatic dishwashing detergents, sodium silicate, sodium bicarbonate and sodium carbonate are the substances with the highest risk index based on the results of the screening. The reason for the high R value is a combination of a relatively low TF and that these substances are used in a higher percentage (see Table 1-2) in the products. The screening of the hazard potential of the substances is based on simple indicators for degradation and aquatic toxicity derived from the DID-list. The screening approach implies that a high value of the risk index (R) does not confirm that the use of the substance in household detergents presents a risk to the aquatic environment, but merely that further assessments are necessary.

The screening points to sodium carbonate and sodium bicarbonate among the substitutes with the highest hazard potential. Carbonate and bicarbonate are prevalent in all natural aqueous systems where they are in equilibrium together. Under strongly alkaline conditions carbonate is predominant, whereas bicarbonate prevails under weakly alkaline conditions. Reservations may be taken to the results from the screening applied here due to the facts mentioned and that the buffer capacity of the receiving aqueous system was not accounted for. Carbonate will be neutralised in wastewater treatment plants and sodium itself has a low toxicity (HERA, 2009).

The following sections 2.3-2.5 include a detailed aquatic risk assessment of the substances characterised by high risk indexes in the screening assessment. The risk assessment will not be done on sodium bicarbonate and sodium carbonate due to the reasons mentioned above.

Table 2-4 Substitutes for phosphate according to their degree of risk (R)

The Swedish report “Fosfater i tvätt- och rengöringsmedel” (Swedish Chemicals Agency, 2006) describes the zeolites as having low toxicity towards aquatic organisms (only minor effects on algae are observed) and it is mentioned that there are no signs of risks to human health (CMR). According to the report, and numerous other references, citrates are easily biodegradable in the environment, they are not bioaccumulative and they are non-toxic to aquatic organisms. Likewise, polycarboxylates have a low toxicity and are not bioaccumulative because of their large molecular weight. Phosphonates are described as having a low biodegradability and as non-bioaccumulative because they are very water soluble. Some phosphonates are reported to have a high toxicity and care should be taken, keeping in mind the low biodegradability.

In this study, an aquatic risk assessment will be conducted for zeolite, sodium silicate, phosphonates and IDS with focus on their use as substitute for phosphates in laundry detergents.

2.3 Derivation of PNEC

The information on the intrinsic ecotoxicological properties of the substances listed in Appendix B was obtained from a review of existing databases (US-EPA, IUCLID and QSAR calculations) as well as an environmental risk assessment for several of the above mentioned substances which was previously conducted by DHI (Danish EPA, 2001). The data on ecotoxicological properties are used for the exact substances where identified by the CAS numbers. For IDS ecotoxicological data were obtained from data on the commercial raw material Baypure^TM CX 100 solid consisting of more than 65 % by weight of IDS.

Values for the aquatic toxicity of substances listed in the Appendix B are given by the results from the studies with the highest toxicity, i.e. studies which have resulted in a low NOEC/EC₅₀/LC₅₀, in order to calculate a Predicted No Effect Concentration (PNEC_aquatic). These values are highlighted in Appendix B.

The derivation of a Predicted No Effect Concentration (PNEC_aquatic) for a substance involves the application of an assessment factor (AF). The assessment factor expresses the difference between the effect concentration values derived from laboratory tests and the derivation of the PNEC_aquatic. The size of the AF depends on the number relevant studies, the number of trophic levels and taxonomic groups covered by the data, and the availability of data from long-term chronic tests. AFs are applied according to the principles listed in the Technical Guidance Document (TGD, 2003) and listed in Table 2-5 together with the derived PNEC_aquatic. For Sodium silicate ecotoxicological data for fish and crustacean show a low toxicity (EC₅₀/LC₅₀/NOEC >100 mg/L) and we would expect the toxicity for algae to be low as well. Therefore, an assessment factor of 1,000 is applied for this compound. An assessment factor of 100 was applied on the long term NOEC for algae, which is considered to be the most sensitive organism.

The PNEC_aquatic in Table 2-5 is representing the chronic PNEC. In general, it may be assumed that the acute PNEC is 10 times higher than the chronic PNEC.

Table 2-5 Calculated PNEC for the aquatic environment

2.4 Derivation of PEC

For the derivation of a Predicted Environmental Concentration (PEC) for the identified substances with highest risk two methods where applied: EUSES and ECO Lab.

2.4.1 Use of EUSES

EUSES is a decision-support instrument, which enables a rapid and efficient assessment of the general risks posed by substances to man and the environment. The EUSES estimation is based on the principles described in the TGD (2003). In this project, EUSES was used to estimate the environmental concentrations of the four phosphate substitutes. For the estimations of PEC, the EU TGD 2003 Risk Assessment Spreadsheet Model (2007) was used. The spreadsheet model was chosen in order to ease the estimations of the PEC values, as the properties of the substances identified in this project allowed the spreadsheet to be used. The two inorganic substances and the phosphonates are considered as conservative substances with no evaporation, absorption or degradation in the sewage treatment plant. Hence, the EUSES calculations represent a worst-case scenario with no treatment. Data on (1-hydroxyethylidene) bis phosphonic acid (CAS nr. 2809-21-4) are used for the estimation of PEC on phospohonates. For IDS the biodegradability of the substance is taken into account in the estimation of PEC.

For each of the four substances, data on physical-chemical substance properties, degradation and transformation rates and emissions are put into the risk assessment spreadsheet.

In order to estimate the local emission to wastewater, the number of households in Denmark has to be taken into account. Information of the number of households given in Table 2-6 is collected from Statistics Denmark (2007).

Table 2-6 Information on numbers of households in Denmark

To estimate the emission to the wastewater, it is assumed that 90% of the waste water from all households is discharged to the sewage system. It is assumed that single families are doing 150 laundries per year and other families are doing 400 laundries per year. The approximations listed in Appendix A regarding use of phosphate-free laundry detergents and the typical concentrations listed in Table 1-2 are used to calculate the emission of each substance. Table 2-7 shows the data used in the EUSES estimation.

Table 2-7 Overview of data used in the EUSES estimation

¹⁾ Total emission of laundry detergents to wastewater in Denmark was estimated to 46,576 t/year

The input values used in the calculations and the output from the calculation are shown in Appendix D. The PEC-values are estimated for the local surface water and the local marine water (Table 2-8). The maximum and the average value are the same using the EUSES model for conservative substances.

Table 2-8 Predicted Environmental Concentration from the EUSES estimation

2.4.2 Emission/inventory and ECO Lab model

In the ECO Lab modelling, Little Belt was chosen as the Danish surface water scenario. Appendix C describes the background for the ECO Lab model in detail. Results are given in Table 2-9.

Table 2-9 Calculated PEC value given as maximum and average values and presented as 90^th percentile, when the concentration in the outlet from the sewage treatment plant is 1 mg/L

PEC90^th : the 90^th percentile

It is assumed that a person uses 200 L of water each day. And that a household of 4 persons has in average 1 wash per day (i.e. 0.25 wash per person per day). Again, the dosage for laundry detergent (X) is assumed to be 70 g/wash.

PEC_inlet = (X (g) × conc. of substance in the detergent (%) × 0.25) / 200 L/day

PEC_outlet = PEC_inlet × (DF) (DF taken from Table 2-1)

PEC_average = PEC _outlet × 90^thPEC_average /1 mg/L

PEC_max = PEC _outlet × 90^th PEC_max /1 mg/L

Where

PEC_inlet is the Predicted environmental concentration in the inlet

PEC_outlet is the Predicted environmental concentration in the outlet of the sewage treatment system

PEC_average is the average Predicted environmental concentration close to the outlet of a STP

PEC_max is the maximum Predicted environmental concentration close to the outlet of a STP

Example for Zeolite (laundry detergent)

PEC _inlet = (70 g/wash × 25 % × 0.25 wash/day)/200 L/day = 0.022 g/L = 22 mg/L

PEC _outlet = 0.022 g/L × 1 = 0.022 g/L = 22 mg/L

PEC_max = 0.1017 × 22 mg/L = 2.2 mg/L

PEC_average = 0.01885 × 22 mg/L = 0.41 mg/L

Table 2-10 Predicted Environmental Concentration from the ECO Lab modelling

2.5 Aquatic risk assessment

In order to evaluate the risk involved with the usage of the high risk substances, a Risk Quotient (RQ) is calculated as the ratio

RQ = PEC/PNEC

The estimation of RQ based on the PEC values obtained by use of the EUSES is shown in Table 2-11. The PEC_local.water is used for estimation of the RQ_acute and the RQ_chronic is derived by the PEC_{local.water,marine}.

Table 2-11 Calculated Risk Quotient (R) based on PEC estimated by use of EUSES

The estimation of RQ based on the ECO Lab modelling is shown in Table 2-10. PEC_average is used to evaluate the RQ of chronic exposure and the PEC_max is used to evaluate the RQ of acute exposure. We earlier derived a PNEC_aquatic which in this case is representing PNEC_chronic (Table 2-5) and it is assumed that the PNEC_acute is 10 times higher than the PNEC_chronic i.e. values in Table 2-5 are multiplied by 10 in order to calculate RQ_acute.

Table 2-12 Calculated Risk Quotient (R) based on PEC derived from the ECO Lab modelling

According to the TGD, if the RQ is below 1, the risk to the environmental compartment is considered to be acceptable. An RQ above 1 indicates a potential risk. Following this, none of the phosphate substitutes pose a risk to the aquatic environment since all calculated RQ values are below 1.

2.6 Conclusions

The environmental risk assessment was carried out for the substances with the highest risk index derived in the screening assessment.The results of the risk assessment indicate that substitution of phosphate in household detergents will lead to environmental concentrations of the substitutes which are below the no observed effect concentrations for the substances. On this basis, it is unlikely that a ban of phosphate in laundry detergents will lead to toxic effects in the aquatic environment. The screening and the risk assessment is based on prototypes for phosphate-free detergents for laundry which are typical for the use today. However, this may be changed by a ban of phosphate, which potentially may lower the price on the new alternative substitutes as IDS, GLDA and MGDA. The substitutes are all considered as easily biodegradable and therefore anticipated to be biodegraded during treatment in the wastewater treatment plants.

| Front page | | Contents | | Previous | | Next | | Top |

Version 1.0 December 2009, © Danish Environmental Protection Agency