Elsevier

Environmental Pollution

Volume 159, Issue 12, December 2011, Pages 3439-3445
Environmental Pollution

Engineered nanomaterials in rivers – Exposure scenarios for Switzerland at high spatial and temporal resolution

https://doi.org/10.1016/j.envpol.2011.08.023Get rights and content

Abstract

Probabilistic material flow analysis and graph theory were combined to calculate predicted environmental concentrations (PECs) of engineered nanomaterials (ENMs) in Swiss rivers: 543 river sections were used to assess the geographical variability of nano-TiO2, nano-ZnO and nano-Ag, and flow measurements over a 20-year period at 21 locations served to evaluate temporal variation. A conservative scenario assuming no ENM removal and an optimistic scenario covering complete ENM transformation/deposition were considered. ENM concentrations varied by a factor 5 due to uncertain ENM emissions (15%–85% quantiles of ENM emissions) and up to a factor of 10 due to temporal river flow variations (15%–85% quantiles of flow). The results indicate highly variable local PECs and a location- and time-dependent risk evaluation. Nano-TiO2 median PECs ranged from 11 to 1′623 ng L−1 (conservative scenario) and from 2 to 1′618 ng L−1 (optimistic scenario). The equivalent values for nano-ZnO and nano-Ag were by factors of 14 and 240 smaller.

Highlights

► We combine probabilistic material flow analysis and graph theory. ► First local concentrations for engineered nanomaterial in rivers are modeled. ► Geographical and temporal flow variation is the most influential parameter. ► Risks to aquatic organisms in urban river sections cannot be excluded.

Introduction

Production and application quantities of engineered nanomaterials (ENMs) are growing and it has to be expected that ecosystems will be exposed to significant levels of such materials (Wiesner et al., 2006, Wiesner et al., 2009, Nowack and Bucheli, 2007, Alvarez et al., 2009, Battin et al., 2009). Scientific awareness has been particularly focussed on aquatic pollution (Boxall et al., 2007, Battin et al., 2009, Farre et al., 2009, Perez et al., 2009). Battin et al. (2009) showed that nano-TiO2 at ambient UV radiation levels and realistic nano-TiO2 concentrations can have relevant impacts on natural microbial aquatic communities. ENMs are also envisaged for applications in environmental remediation techniques (Vaseashta et al., 2007, Baun et al., 2008) that lead to direct release of such material into aquifers (Zhang, 2003). Industrial and domestic products and wastes containing ENMs tend to end up in water: either they are released directly into rivers and lakes, e.g. from outdoor use of sunscreens, or indirectly via surface run-off, domestic or industrial wastewater (Moore, 2006).

Unfortunately, the quantitative detection of ENMs in the environment and the distinction between engineered and naturally occurring nanomaterial is still extremely limited (Tiede et al., 2008, Hassellöv and Kaegi, 2009). Thus, modeling the ENMs’ release to and fate in the environment is essential to estimate environmental exposure. Such estimations have to cover diffuse emissions from a large number of relevant ENM containing products and lifecycle stages. These include but are not limited to ENM release into the environment from ENM production, ENM incorporation into products and storage, use, waste generation and disposal of such products. Once released into the environment, this material will to some extent agglomerate, associate with suspended solids or sediments, potentially accumulate in organisms and enter the food chain or drinking water sources (Boxall et al., 2007).

Mass balance partitioning models have already been used to assess aquatic exposure to ENM. Boxall et al. (2008) proposed simplistic algorithms to predict concentrations of ENMs in water, soil, and air by means of a specified but non-comprehensive range of ENM applications. Blaser et al. (2008) calculated the flow of silver in the environment including dissolved silver released from nano-Ag in textiles and biocidal plastics. However, biocidal plastics and textiles were seen to account only for up to 15% of the total silver release into water. Water and sediment concentrations were estimated based on a river box model developed by Scheringer et al. (1999), where each box was subdivided into a compartment of moving water, a compartment of stagnant water and a sediment compartment. Recently, ENM release and concentrations in the environment were predicted by analyzing the complete lifecycle of ENM and ENM containing products (Mueller and Nowack, 2008, Gottschalk et al., 2009). In these studies predicted environmental concentrations (PECs) and predicted no effect concentrations (PNECs) revealed risk quotients (PEC/PNEC) at regional (national) level indicating that at least risks to aquatic organisms from nano-TiO2, nano-Ag and nano-ZnO in undiluted sewage treatment plant (STP) effluents cannot be excluded. O’Brien and Cummins (2010) predicted relative concentrations in surface water for environmental release of ENM from selected single products. Arvidsson et al. (2011) modeled particle number concentrations in aquatic environments by considering continuous ENM release and by focusing on possible particle agglomeration kinetics.

In order to overcome one of the main limitations of such PEC calculations at regional scale – i.e. the assumption of homogeneous material distribution within aggregated environmental compartments – the modeling of concentrations at higher spatial and temporal resolutions at a local level is required (ECB, 2003). As a consequence, to evaluate local, specific ENM concentrations in rivers, the following refinements are necessary: (i) the total (national) ENM emission should be distributed geographically, e.g. according to spatial population density; (ii) geographical and temporal variation of flow rates in rivers have to be considered; and (iii) the transport and fate of ENMs within the system – i.e. the real river network – need to be studied in more detail. The modeling framework presented in this study comprehensively addresses all three points mentioned above. The systems analysis enables a clear distinction between model input uncertainty (ENM emission and transport in rivers) and spatial and temporal variability, which is expected to be influenced by the geographical position in the river network.

Since previous results predicted the highest concentrations and risk coefficients in surface water for nano-TiO2, nano-Ag, and nano-ZnO (Gottschalk et al., 2009), our study aims to calculate local PECs for these materials. Such concentrations are modeled along all rivers in Switzerland at base flow. This enables us to identify river sections where PECs are expected to exceed PNECs at minimum dilution. Additionally, for selected river sections complete hydrological information will allow us to expand the geographic evaluation by including temporal differentiation in the estimation of PECs and potential risks for aquatic organisms exposed to ENMs.

Section snippets

ENM emission

The total annual and regional (national) amount of ENM (nano-ZnO, nano-Ag) discharged to rivers was taken from Gottschalk et al. (2009). The emission of nano-TiO2 was computed by using new data on ENM production volumes (Robichaud et al., 2009) and the same lifecycle-based Monte Carlo (MC) method (Gottschalk et al., 2010). The modeled emissions reflect direct release of ENM during the whole lifecycle of these compounds and the products containing them (consumption and disposal phases), indirect

Spatial variation of nano-TiO2 concentrations at Q95%

The modeled PECs depend on the geographical distribution of the STP discharges, the local dilution in rivers and the transport scenario. The evaluation of the spatial distribution of nano-TiO2 is shown in Fig. 1. The highest concentrations were found in the midlands or near urban centers. The calculated values in rural, alpine and pre-alpine areas were negligibly small. Exceptions are river sections in the area of large tourist destinations in the Alps, where the PECs were comparable to the

Discussion

In order to model realistic PECs of ENM in rivers it is essential to i) account for realistic local dilution (data available, but not considered in fate models), ii) reduce uncertainty on ENM release (gather more detailed input data independent of model complexity: geographically resolved information on ENM production and use as well as empirical information on the ENM release from products containing these materials) and iii) consider the fate and behavior of these compounds in the liquid

Conclusions

This study differentiates for the first time between uncertainty of input parameters and effects of measurable, natural variability (for ENM and calculations at high spatial and temporal resolution). Geographical variation – caused by the local distribution of ENM emissions and spatially variable, local dilution – has by far the greatest influence on the modeled PECs revealing a factor of up to 300 (15%–85% quantiles) for the optimistic scenario with rapid ENM removal from water. In comparison,

Acknowledgments

This work was supported by internal Empa funding and the Swiss Federal Office for the Environment (FOEN). We thank Mario Keusen (FOEN) for creating the maps.

References (34)

  • A. Baun et al.

    Ecotoxicity of engineered nanoparticles to aquatic invertebrates: a brief review and recommendations for future toxicity testing

    Ecotoxicology

    (2008)
  • N.T. Boncagni et al.

    Exchange of TiO2 nanoparticles between Streams and Streambeds

    Environmental Science & Technology

    (2009)
  • A.B. Boxall et al.

    Engineered nanomaterials in soils and water: how do they behave and could they pose a risk to human health?

    Nanomedicine

    (2007)
  • A.B.A. Boxall et al.

    Current and Future Predicted Environmental Exposure to Engineered Nanoparticles

    (2008)
  • Technical Guidance Document on Risk Assessment

    (2003)
  • M. Farre et al.

    Ecotoxicity and analysis of nanomaterials in the aquatic environment

    Analytical and Bioanalytical Chemistry

    (2009)
  • F. Gottschalk et al.

    Modeled environmental concentrations of engineered nanomaterials (TiO2, ZnO, Ag, CNT, fullerenes) for different regions

    Environmental Science and Technololgy

    (2009)
  • Cited by (140)

    View all citing articles on Scopus
    View full text