MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater

Miljø- og Fødevareudvalget 2016-17
MOF Alm.del Bilag 303
Offentligt

Microplastic in Danish

wastewater

Sources, occurrences

and fate

Environmental Project

No. 1906

December 2016

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Publisher: The Danish Environmental Protection Agency

Editor:

Professor Jes Vollertsen, Aalborg University and

Aviaja Anna Hansen, Krüger A/S

ISBN: 978-87-93529-44-1

The Danish Environmental Protection Agency publishes reports and papers about research and development projects

within the environmental sector, financed by the Agency. The contents of this publication do not necessarily represent

the official views of the Danish Environmental Protection Agency. By publishing this report, the Danish Environmental

Protection Agency expresses that the content represents an important contribution to the related discourse on Danish

environmental policy.

Sources must be acknowledged.

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Contents

Preface 6

Background of the study

Steering group

Advisory group

Project group

Conclusions and summary

Konklusion og sammenfatning

1.1

1.1.1

1.2

1.2.1

2.1

2.1.1

2.2

3.1

3.2

3.3

4.1

4.1.1

4.1.2

4.1.3

4.1.4

4.1.5

4.1.6

4.2

4.2.1

Introduction

Microplastic in wastewater and wastewater treatment plants

Fate of microplastic in wastewater treatment plants

Objectives of the study

Definition of microplastic in the present study

Project design

Occurrences, sources and fate of microplastic in Danish wastewater treatment plants

The wastewater treatment plants investigated

Occurrences and sources of microplastic on Danish farmlands

Methodology

Choosing the analytical method

The methodology at a glance

Method validation

Results and discussion

Microplastic in Danish wastewater treatment plants

Microplastic concentrations in raw and treated wastewater

Size distributions of microplastic particles

Microplastic concentrations in sludge

Mass balance of microplastic in wastewater treatment plants

Microplastic loads on the aquatic environment

The polymer composition of the microplastic particles

Occurrences and sources of microplastic on Danish farmlands

The impact of wastewater sludge on agricultural soil

Perspectives

Analytical method

Sampling

Acknowledgement

References

Appendix 1.

Appendix 1.1

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Appendix 1.2

Appendix 1.3

Appendix 1.4

Appendix 2.

Appendix 2.1

Appendix 2.2

Appendix 2.3

Appendix 2.4

Appendix 2.5

Analysis of microplastic in wastewater and soil

Spectral analysis

Calculation of particle volume and mass

Method validation, uncertainties, and detection limits

Recovery of plastic particles

Uncertainties

Detection limits

Identification of tire rubber

Data overview

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Preface

Background of the study

Microplastic in the environment was for the first time described by marine biologists in 2004

(Thompson et al, 2004) and research of microplastic has until recently mainly been driven by

the field of marine biology. Microplastic emissions to the environment have in the resent years

gained increasing political awareness, where also sources and reduction potentials have been

on the agenda.

In Denmark a report on microplastic occurrence, effects and sources was published by the

Danish Environmental Protection Agency in 2015, where wastewater treatment plants

(WWTPs) were identified as potential important sources of microplastic emission to the marine

waters of Denmark (Lassen et al., 2015).

To follow up on this survey the present study was initiated by the Danish Environmental Protec-

tion Agency to elucidate the role of WWTPs in the microplastic emissions to the environment.

This project was part of the Danish Government initiative to improve the understanding about

sources and effects and the possibilities to reduce microplastic pollution in the environment.

Funding for this activity were allocated on the Finance Act for 2015-2016.

Steering group

The steering group of the project consisted of:

Flemming Ingerslev, Danish Environmental Protection Agency

Linda Bagge, Danish Environmental Protection Agency

Jes Vollertsen, Aalborg University

Vibeke Borregaard, Krüger A/S

Aviaja Anna Hansen, Krüger A/S

Advisory group

An advisory group with representation of various experts in the field of wastewater and micro-

plastics has followed the project:

Per Helmgaard, Danish Nature Agency

Rikke Joo Vienberg, Danish Nature Agency

Henrik Andersen, Technical University of Denmark

Nanna Hartmann, Technical University of Denmark

Annemette Palmquist, Roskilde University

Hanne Løkkegaard, Danish Technological Institute

Dines Thornberg, BIOFOS

Project group

The experimental analysis has been conducted by the research group of Professor Jes Vol-

lertsen at Aalborg University:

Marta Simon, Aalborg University

Nikki Van Alst, Aalborg University

Diana A. Stephansen, Aalborg University

Amelia Borregaard, Aalborg University

Kristina Borg Olesen, Aalborg University

Sampling was conducted by Anders Lund, Krüger A/S in close collaboration with the skilled

operation personnel at the investigated WWTPs.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Conclusions and summary

The objectives of the present study were to evaluate the role of Danish wastewater treatment

plants (WWTPs) in the emission of microplastic to the environment in terms of amounts and

types of plastic polymers emitted and if possible, to evaluate which sources these plastic poly-

mers could originate from.

Samples from 10 WWTPs (wastewater, inlet and outlet), sludge from 5 of these plants, and 10

farmlands soils (5 soils that had received sludge as fertilizer and 5 that had not) were analysed

for the occurrences of microplastic in the size range 20-500 µm with the currently most ad-

vanced method available for microplastic investigations (Fourier Transformed Infrared Spec-

troscopy imaging applying a Focal Plane Array). This method allows both determination of the

microplastic concentrations in the samples and identification of the type of plastic polymer of

each microplastic particle.

The investigation was designed as a general screening study of Danish wastewater and the

results are therefore an estimation of the occurrences of microplastic in average Danish

wastewater, thus the results are indicative for the overall Danish wastewater, but not the distinct

WWTPs.

Microplastic concentrations in wastewater and emission from WWTPs

In the raw wastewater the microplastic concentration was quantified to a median of 1.3 10

particles/L (size range 20-500 µm) corresponding to 5.9 mg/L, which is equivalent to one per-

cent of the total organic matter of the raw wastewater, as it typically holds 320-740 mg COD/L.

In treated wastewater the microplastic concentration was quantified to a median of 5,800 parti-

cles/L (size range 20-500 µm) corresponding to 0.02 mg/L. The variability of microplastic con-

centrations in raw wastewater between the 10 investigated treatment plants was quite large

ranging from 13,000 to 442,000 particles/L corresponding to 0.2 to 30 mg/L.

The average emission from a Danish WWTP to the aquatic environment is from this calculated

to 0.3% (with 25 and 75 percentiles of 0.0% and 0.7%) of the microplastic mass coming into

the plant.

From the results obtained from the analysis of the wastewater samples it is thereby shown that

the emission of microplastic from Danish WWTPs to the receiving waters is minor compared to

the total load on the plants.

Microplastic concentrations in sludge

The median microplastic concentration in the wastewater sludge was quantified to 169,000

particles/g dewatered sludge corresponding to 4.5 mg/g dewatered sludge, which means that

approximately 0.7% of the dewatered sludge was microplastic.

Mass balance of microplastic in wastewater treatment plants and emission to the aquatic

environment

A rough mass balance can be made on the inlet and outlet mass of microplastic in wastewater.

Assuming that microplastic is inert in the treatment plant, the total mass in the inlet must equal

the sum of the mass in the sludge and in the discharged wastewater.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Based on the results, it is estimated that the annual load of microplastic to all Danish treatment

plants is 4,000 ton/year (1,124 - 5,072 ton/y, 25 and 75 percentiles), where 11 ton/year (4.9 -

16 ton/y, 25 and 75 percentiles) is discharged with the treated wastewater and the remaining

fraction measured to 3,100 ton/y (with 25 and 75 percentiles of 970 and 3,110 ton/y) is incor-

porated into the sludge (Figure 1). Approximately ¾ of the total microplastic mass load on the

treatment plants are from the obtained concentrations accounted for. The lacking ¼ of the mi-

croplastic mass may simply be due to measurement uncertainties or other unresolved dynamics

in the WWTP e.g. degradation of certain polymers.

Mass balance of microplastic in Danish wastewater

Inlet

wastewater

1,100 - 5,100 t/y

WWTP

5 - 16 t/y

Emission to the

aquatic environment

1,000 - 3,100 t/y

Incorporated into

the sludge fraction

Figure 1. Mass balance of microplastic (size range 20-500 µm) in Danish wastewater. The

rounded numbers of 25 and 75 percentiles are shown.

Assuming a total microplastic load to the Danish aquatic environment of 600-3,100 ton/year

(size range 1µm - 5mm) as estimated by Lassen et al. (2015), the emission from the WWTPs to

the aquatic environment of 5 -16 ton/y (size range 20-500 µm) represents as a worst case 3%

of the total emitted microplastic to the Danish aquatic environment. It can therefore be

concluded that discharge of treated wastewater from the municipal treatment plants has a minor

role in terms of microplastic emission to the aquatic environment and that other sources such

as stormwater runoff, combined sewer overflows, atmospheric deposition, and etcetera likely

are more important sources.

Microplastic concentrations in agricultural soils

The concentration of microplastic in the soils was low and accounted between 0,0001 and

0,001% (w/w) of the soil. The median microplastic concentration in the investigated agricultural

soils was 5.8 mg/kg soil (with 25 and 75 percentiles of 1.4 and 7.6 mg/kg soil), when sludge

had been added as fertilizer and 12 mg/kg soil (with 25 and 75 percentiles of 4.4 and 14.9

mg/kg soil), when no sludge had been added to the soils. This means that higher concentra-

tions of microplastic were found in soils where sludge had not been added as fertilizer. Consid-

ering the role of the sludge fertilizer in microplastic emission to agricultural soils it is estimated

from the obtained results that sludge fertilization will increase the microplastic concentration of

the soil by approximately 15 mg/kg soil (6.7-22 mg/kg, 25 and 75 percentiles), when assum-

ing a tilling depth of 30 cm (i.e. the soil depth into which the sludge is mixed into). This indicates

that sludge is just one of many sources of microplastic emission to the agricultural soils and

further investigations are needed to understand the importance of various microplastic sources

for accumulation in farmland soils e.g. windborne litter could be an important source.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

The polymer types of the microplastic particles in wastewater, sludge and soil

The by far most abundant plastic material in the wastewater samples, both inlet and outlet, was

polyamide/nylon, maybe originating from various forms of textiles, clothing and carpets. Poly-

ethylene and co-polymers and zinc stearate coated particles were also detected in significant

amounts, while polypropylene and PVC were detected in smaller proportions. The distribution of

the different plastic polymers were more or less the same in the inlet and the outlet wastewater

meaning that the treatment plant does not to any significant extent preferentially remove specif-

ic plastic polymers. Interestingly, the distribution of plastic polymers was different in the sludge-

fraction, where polyethylene was the dominant plastic material followed by polyamide/nylon and

polypropylene. This discrepancy between the wastewater and sludge could indicate that the

anaerobic digestion process affects the plastic, either by breaking it down to particles too small

to detect by the applied approach or by biological degradation. The latter is known to be possi-

ble for polyamide, but further investigations are needed to understand how and where these

changes are occurring in the sludge and whether it is a matter of random variability as a conse-

quence of the screening approach applied in the present study.

The dominant plastic material in soils that had not received sludge was polyethylene followed

by polyamide/nylon and polypropylene, and it had more or less the same distribution as ob-

served in the sludge. Polypropylene, the polymer only observed in relatively low abundances in

the wastewater and sludge samples, was found to be the dominant polymer in the soils that had

received sludge. This could indicate that polypropylene is more withstanding to disruption and

degradation, but more samples need to be investigated to exclude random variability due to the

small sample size investigated in the present screening study. Both polyethylene and polypro-

pylene originate from a wide pallet of products, including packaging materials such as plastic

bags, plastic films, plastic bottles, and so on. The study did not reveal a single rubber particle

from tire abrasion (styrene butadiene co-polymers) even though tire abrasion is identified as the

largest microplastic source released to the Danish environment (Lassen et al., 2015). The most

likely reason is that such particles were smaller than the 20 μm, which were the lower size limit

for detection in the present study.

Methods for evaluation of microplastic in wastewater, sludge and soil samples

The observations made in this study lead to the conclusion that when addressing the efficiency

of wastewater treatment plants to retain microplastic, mass as the unit of measurement is signif-

icantly more reproducible than particle numbers. The number of particles is affected by physical

breakdown processes, and this breakdown can result in increases in particle numbers without

an increase in plastic mass. Hence, when applying only particle numbers for quantifying the

efficiency of a treatment system, this system could in principle ‘produce’ microplastic because

larger particles were broken down into smaller particles. On the other hand, when it comes to

the impact of microplastic on aquatic fauna, the number of particles plays potentially a signifi-

cant role. Hence microplastic mass should be used to assess treatment efficiencies and particle

numbers should additionally be reported to support environmental impact assessment.

With the experience from the method development and optimisation of a valid approach for

microplastic identification and quantification carried out in the present study, it is evident that

microplastic results reported in literature should be carefully reviewed and the method used for

detection of microplastic should be looked over before acknowledging the reported results.

Light microscopy alone is unsuitable for investigations of microplastic in environmental samples

and verification of the particle material as plastic is highly important. Therefore, we recommend

that future investigations of microplastic in environmental samples should be conducted with

either FT-IR or Raman spectroscopy methods. We further recommend that sampling methods

and analytical methods for microplastic analysis are standardized to allow comparison between

results of microplastic monitoring and investigations.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Konklusion og sammenfatning

Formålet med nærværende studie har været at evaluere danske renseanlægs rolle i udlednin-

gen af mikroplast til miljøet både i henhold til udledningsmængder og hvilke typer af plastikpo-

lymerer, der udledes og hvis muligt, hvilke kilder disse plastikpolymerer kan stamme fra.

Prøver fra 10 renseanlæg (indløbs- og renset spildevand), slam fra 5 af disse renseanlæg og

10 landbrugsjorde (5 jorde som har fået tilført slam som gødskning og 5 jorde, som ikke har fået

tilført slam) er blevet analyseret for indholdet af mikroplast i størrelsesordenen 20-500 µm med

den mest avancerede metode til måling af mikroplast, der eksisterer i dag (Fourier Transforme-

ret Infrarød Spektroskopi billedbehandling med Focal Plane Array, FT-IR). Denne metode mu-

liggør både bestemmelsen af mikroplastkoncentrationerne i prøverne og identifikation af, hvilke

plastpolymerer mikroplastpartiklerne består af.

Studiet er designet som et screeningsstudie af dansk spildevand og resultaterne er derfor et

estimat af indholdet af mikroplast i dansk gennemsnits spildevand og indikationer for dansk

spildevand generelt og dermed ikke spildevand fra specifikke renseanlæg.

Mikroplastkoncentrationer i spildevand og udledningen fra renseanlæg

I indløbsspildevandet blev medianen af mikroplastkoncentrationen estimeret til 1,3 10 partik-

ler/L (størrelsesordenen 20-500 µm) svarende til 5,9 mg/L. Dette udgør en procent af totalind-

holdet af organisk materiale i indløbsspildevand, som typisk er i størrelsesordenen 320-740 mg

COD/L. I renset spildevand var medianen af mikroplastkoncentrationen 5.800 partikler/L (stør-

relsesordenen 20-500 µm) svarende til 0,02 mg/L. Variationen i mikroplastkoncentrationen i

indløbsspildevandet mellem de 10 undersøgte renseanlæg var relativ stor og lå mellem 13.000

og 442.000 partikler/L svarende til 0,2 og 30 mg/L.

Den gennemsnitlige udledning af mikroplast til vandmiljøet fra et dansk renseanlæg er ud fra

ovenstående beregnet til 0,3% (med 25 og 75% fraktiler på 0.0% og 0.7%) af massen af mikro-

plast, som kommer ind på renseanlægget.

Fra de opnåede resultater fra analysen af spildevandsprøverne er det dermed vist, at udlednin-

gen af mikroplast fra danske renseanlæg til vandmiljøet er lav i forhold til de mængder som

ledes til renseanlæggene.

Mikroplastkoncentrationer i slam

Medianen af mikroplastkoncentrationen i spildevandsslammet blev kvantificeret til 169.000

partikler/g afvandet slam svarende til 4,5 mg/g afvandet slam, hvilket betyder at omkring 0,7%

af det afvandede slam var mikroplast.

Massebalance for mikroplast på danske renseanlæg og udledningen til vandmiljøet

Hvis det antages, at mikroplast er inert på renseanlæg, så den totale masse af mikroplast, der

kommer ind på renseanlæg er lig med summen af massen i renset spildevand og massen i

spildevandsslammet. Dermed kan en grov massebalance opstilles for dansk spildevand.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Udfra de opnåede resultater er det estimeret, at den årlige tilførsel af mikroplast til alle danske

renseanlæg er 4.000 tons/år (1.124 - 5.072 tons/år, 25 og 75% fraktiler), hvor 11 tons/år (4,9-16

tons/år, 25 og 75% fraktiler) bliver udledt med det rensede spildevand og den resterede fraktion

målt til 3.100 tons/år (25 and 75% fraktiler på 970 og 3,110 tons/år) bliver indbygget i slamfrak-

tionen (Figur 2). Omkring ¾ af den totale masse af mikroplast som kommer ind på renseanlæg-

gene er dermed gjort rede for. Den resterede ¼ af massen af mikroplast kan enten skyldes

måleusikkerheder eller andre uafklarede dynamikker i renseanlæggene, der påvirker mikropal-

sten eks. nedbrydning af specifikke polymerer.

Massebalance for mikroplast i dansk spildevand

Indløbs

spildevand

1,100 - 5,100 t/år

Renseanlæg

5 - 16 t/år

Udledning til vandmil-

jøet

1,000 - 3,100 t/år

Indbygning i slam-

fraktionen

Figur 2. Massebalance for mikroplast (størrelse 20-500 µm) i dansk spildevand. Tallene er

afrundede værdier af 25 og 75% fraktiler.

Hvis det antages, at den totale udledning af mikroplast til det danske vandmiljø er i størrelsen

600-3.100 tons/år (størrelse 1µm til 5mm) som estimeret af Lassen et al. (2015), så udgør de 5-

16 tons/år (størrelse 20-500 µm) udledt med renset spildevand i værste fald 3% af den totale

udledning af mikroplast til dansk vandmiljø. Det kan derfor konkluderes, at renset spildevand fra

renseanlæg udgør en mindre rolle i udledningen af mikroplast til det danske vandmiljø og at

andre kilder som eksempelvis vejvand, overløb, atmosfærisk deponering og lignende formentlig

er vigtigere kilder.

Mikroplastkoncentrationer i landbrugsjorde

Koncentrationen af mikroplast i de analyserede jorde var lav og udgjorde mellem 0,0001 og

0,001% (m/m) af jorden. Medianen af mikroplastkoncentrationen i de analyserede landbrugs-

jorde var 5,8 mg/kg jord (1,4-7,6 mg/kg jord, 25 og 75% fraktiler) i de jorde, som havde fået

tilført slam som gødskning og 12 mg/kg jord (4,4-14,9 mg/kg jord, 25 og 75% fraktiler) i de jor-

de, hvor der ikke var tilført slam. Dette betyder, at der blev fundet højere koncentrationer af

mikroplast i jorde, som ikke havde fået tilført slam eller andet organisk affald. Ved vurdering af

slamudbringningens rolle i udledning af mikroplast til miljøet er det ud fra resultaterne estimeret,

at slam på landbrugsjord vil øge jordens mikroplastkoncentration med omkring 15 mg/kg jord

(6,7-22 mg/kg, 25 og 75% fraktiler), når det antages, at pløjedybden er 30 cm (den jorddybde

som slammet blandes i). Dette er i samme størrelsesorden som den detekterede plastmængde

i jorde uden tilført slam, hvilket indikerer, at slam kun er én af mange kilder til mikroplastudled-

ning til landbrugsjord. Videre undersøgelser er nødvendige for at opnå forståelse for vigtighe-

den af forskellige mikroplastkilder i akkumuleringen af mikroplast i landbrugsjord eksempelvis er

luftbåren affald måske en vigtig kilde.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Mikroplastpolymertyper i spildevand, slam og jord

Den mest udbredte plastpolymer i spildevandsprøverne, både i indløbs- og renset spildevand,

var polyamid/nylon, som formentlig stammer fra forskellige typer af tekstiler, tøj og gulvtæp-

per. Polyetylen og co-polymerer og zink stearat belagte partikler blev også detekteret i signifi-

kante mængder, mens polypropylen og PVC blev detekteret i mindre mængder. Fordelingen

af de forskellige plastpolymerer var mere eller mindre den samme i indløbs- og renset spilde-

vand, hvilket betyder, at renseanlæg ikke fjerner nogen polymertyper bedre end andre. Forde-

lingen af plastpolymerer var anderledes i slamfraktionen, hvor polyetylen var den mest ud-

bredte polymer efterfulgt af polyamid/nylon og polypropylen. Denne forskel mellem spildevan-

det og slammet kunne indikere, at den anaerobe udrådning påvirker plastpolymererne enten

ved at neddele dem til partikler mindre end 20µm og dermed for små til at blive detekteret i

dette studie eller ved biologisk nedbrydning. Biologisk nedbrydning er vist for polyamid, men

yderligere undersøgelser er nødvendige for at forstå, hvordan og hvor disse ændringer sker i

slammet og om det i stedet skyldes tilfældig variation af screeningsstudiet.

Den dominerende plastpolymer i jord, som ikke havde fået tilført slam var polyetylen efterfulgt

af polyamid/nylon og polypropylen. Denne type jord havde mere eller mindre den samme

udbredelse af polymerer som observeret for slam. Polypropylen blev observeret i lav udbre-

delse i spildevand og slam, men var den dominerende plastpolymer i jord, som havde fået

tilført spildevandsslam som gødskning. Dette kunne indikere, at polypropylen er mere mod-

standsdygtig mod neddeling og nedbrydning end de andre polymerer, men flere analyser og

flere prøver er nødvendige for at udelukke tilfældig variation pga. den lille prøvestørrelse som

er undersøgt i dette screeningsstudie. Både polyetylen og polypropylen indgår i mange plast-

produkter, herunder emballage som plastposer, plastfilm, plastflasker ol. På trods af, at dæk-

afslid er identificeret som den største kilde til mikroplastudledning til miljøet i Danmark (Lassen

et al. 2015) er der i nærværende undersøgelse ikke fundet en eneste gummipartikel fra dæk-

afslid (styren butadien co-polymerer). Dette skyldes højest sandsynligt, at sådanne partikler er

mindre end 20µm, som var den mindste størrelse inkluderet i studiet.

Metoder til evaluering af mikroplast i spildevand, slam og jord

Det kan konkluderes fra observationerne i dette studie, at når effektiviteten af mikroplasttilba-

geholdelsen i renseanlæg evalueres, så er masse som enhed signifikant mere reproducerbar

end partikelantal. Antallet af partikler er påvirket af fysisk neddeling og dette kan føre til flere

partikler uden det øger massen af plast i renseanlægget. Derfor vil evalueringen af rensesy-

stemer med partikelantal kunne konkludere, at der ’produceres’ mikroplastpartikler i systemet

fordi store partikler neddeles til små partikler. Antallet af partikler er dog potentielt vigtig ved

evaluering af effekter på akvatisk fauna. Derfor skal massen af mikroplast anvendes ved eva-

luering af renseeffektiviteter, mens partikelantal ligeledes skal rapporteres til at understøtte

evalueringen af den miljømæssige effekt.

Erfaringerne fra nærværende studies metodeudvikling og metodeoptimering til identifikation

og kvantificering af mikroplast viser, at mikroplastresultater rapporteret i litteraturen skal tilgås

med varsomhed og den anvendte metode brugt til mikroplastundersøgelser skal gennemgås

grundigt før de opnåede resultater bør citeres. Lysmikroskopi som eneste metode er ikke

tilstrækkelig til undersøgelser af mikroplast i miljøprøver og en verificering af, at partiklerne er

plast er vigtigt. Vi anbefaler derfor, at fremtidige undersøgelser af mikroplast i miljøprøver

anvender enten FT-IR eller Raman spektroskopi metoder. Vi anbefaler desuden, at prøvetag-

ning og analytiske metoder til mikroplastanalyser bliver standardiserede således sammenlig-

ninger studier imellem er mulig.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

1. Introduction

1.1

Microplastic in wastewater and wastewater treatment plants

Microplastic is often defined as plastic particles smaller than 5 mm. Microplastic is divided into

primary microplastic and secondary microplastic, where primary microplastic is plastic particles

smaller than 5 mm used in industry or commercial products e.g. in personal care products, raw

materials for plastic production, rubber granules for artificial turfs etc.. Secondary microplastic is

microplastic particles eroded from larger plastic objects e.g. tires, textiles, footwear, paints etc..

In connection with wastewater treatment plants (WWTPs) it is the microplastic which is of main

interest, as plastic fragments larger than 5 mm are removed from the wastewater by the

screens, which are the first step of the wastewater treatment.

The knowledge of microplastic in wastewater is very limited and the studies reported so far

suffer from use of different methodologies, which makes the obtained results more or less in-

comparable (Pedersen and Winther-Nielsen, 2015). The methodological development of tech-

nologies to detect and quantify microplastic has taken a major leap the past years, which has

revolutionised the research field of microplastic. However due to the method novelty, still no

standardized method for investigation of microplastic exists and development of a general pro-

tocol for sample pre-treatment and for identification and quantification of microplastic in envi-

ronmental samples needs to be established.

1.1.1

Fate of microplastic in wastewater treatment plants

Studies of microplastic in wastewater have shown that the majority of the microplastic entering

WWTPs is ending up in the sludge fraction (90-95%; Magnusson and Wahlberg, 2014), while

the fraction emitted to the aquatic environment from the WWTPs is mainly the smaller plastic

particles. Magnusson and Wahlberg (2014) showed that 10-30% of plastic particles in

wastewater in the sizes 20-300 µm were emitted through the outlet to the aquatic environment,

while only 0-1% of the plastic particles larger than 300 µm was emitted to the receiving waters.

Therefore, the flow of microplastic seems to follow the flow of sludge in a WWTP (Figure 3).

Primary sludge

sedimentation

Inlet

Screen

(3-6 mm)

Secondary sludge

sedimentation

Outlet

Sludge

InletInlet

treatment

Figure 3. The flow of microplastic probably follows the flow of wastewater sludge in a

wastewater treatment plant (white arrows).

Incineration

Farmland

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

In Denmark most sludge is used as fertilizer on farmland (775%; Sckerl, 2012), while the rest

is incinerated. This means that most microplastic in wastewater is either combusted when

sludge is incinerated or distributed on farm soils.

1.2

Objectives of the study

The objectives of the study are:

To evaluate the role of Danish WWTPs in the emission of microplastic to the environ-

ment

To evaluate the fate of microplastic entering the Danish WWTPs

To evaluate the fate of microplastic in sludge distributed on Danish farmland

To determine the main types of plastic polymers in the Danish wastewater and from

this give an assessment of the possible sources of the microplastic in Danish

wastewater

The study is designed as a screening study of Danish wastewater in general and the results are

therefore an estimation of the occurrences of microplastic in average Danish wastewater mean-

ing that the results are indicative for the overall Danish wastewater and not distinct WWTPs.

The method used for the microplastic detection is FT-IR imaging (Fourier Transformed Infrared

Spectroscopy applying a Focal Plane Array), which not only allows the representation of the

results in numbers of particles, but also the mass of microplastic and the determination of the

plastic polymer each detected microplastic particle consists of. FT-IR imaging is currently the

most advanced method available for microplastic investigations (Löder and Gerdts, 2015).

A representation of microplastic in terms of both particle quantity and mass has not previously

been reported, but is necessary to allow relation of the results to other studies, for example to

studies on microplastic sources and occurrence in the environment, e.g. Lassen et al. (2015).

This way it becomes possible to evaluate the distribution of microplastic in the sample and to

calculate mass balances, which further allows evaluation of degradation and erosion of the

microplastic in the environment and in biotechnological installations such as WWTPs.

1.2.1

Definition of microplastic in the present study

There is currently no clear accepted definition of what microplastic is, but in literature most

studies define microplastic as plastic fragments from 1µm to 5 mm. Microplastic is in the pre-

sent study defined as polymers of a synthetic material in the size range smaller than 5 mm in all

dimensions. The present screening study has investigated microplastic in the lower size range,

namely 20-500 µm, and has not attempted to quantify microplastic particles above 500 µm.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

2. Project design

2.1

Occurrences, sources and fate of microplastic in Danish

wastewater treatment plants

A screening approach has been applied and the project was designed to evaluate as much of

the Danish wastewater as possible in the given time frame. This was realised by investigating

the wastewater of ten of the largest Danish WWTPs, thereby giving an estimation of the occur-

rence of microplastic in average Danish wastewater. The project has not been designed to give

exact occurrences of microplastic at a given WWTP or to evaluate the variation of microplastic

occurrences between specific WWTPs. Therefore, the obtained results are presented so that

they cannot be traced back to the WWTP of their origin.

2.1.1

The wastewater treatment plants investigated

Wastewater and sludge samples were collected from ten of the largest WWTPs representing

26% of the total Danish wastewater volume (Table 1). Samples were collected at dry weather

and the maximum of rain was 3 mm for 48 h before and during sampling (72 h in total). The dry

weather criterion ensures that the samples are comparable. The occurrences of microplastic

was analysed for the inlet and outlet wastewater from the 10 WWTP (20 samples in total) and

occurrence of microplastic in sludge was analysed for 5 of these plants (five samples in total).

Table 1. Wastewater treatment plants (WWTP) included in the study and the volume of

wastewater treated at each plant. Water volumes are from Miljøministeriet (2015).

WWTP

Water volume

(1000 m /y)

55.044

23.058

19.426

18.608

9.319

9.197

9.032

8.651

8.340

7.563

168.238

Fraction of total Danish

wastewater

Lynetten

Damhusåen

Ejby Mølle

Aalborg Vest

Marselisborg

Herning

Vejle

Kolding

Fredericia

Horsens

Total

26%

2.2

Occurrences and sources of microplastic on Danish

farmlands

The occurrences of microplastic on Danish farmlands were investigated by analysing five fields

that never have received sludge as fertilizer and five fields that have received sludge as fertiliz-

er within the past few years. From this the role of sludge in the microplastic emission to the soil

environment was evaluated.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

3. Methodology

Research of microplastic in the environment is an emerging field, and most of the research that

has been done on this topic has been related to the marine environment. Often samples have

been collected by dragging algae nets behind boats or collecting sand samples from beaches.

Collected samples have in most cases been characterized visually by size and colour of the

catch. This approach is reasonable for particles above, approximately 0.5 mm and especially

where there are few other organic particles than plastic in the sample. However, even for such

particles, visual inspection does not allow a characterization of plastic polymers.

3.1

Choosing the analytical method

The studies that have addressed smaller particle sizes and at the same time identified the plas-

tic polymer are limited. The present study addresses the analytical quantification of 20-500 μm

microplastic particles in raw wastewater, treated wastewater, wastewater sludge, and agricul-

tural soils, and for these matrixes only one or two studies have applied comparable technolo-

gies. In order to study these sample types reliably, a method for sample preparation, sample

concentration, and FT-IR imaging had to be developed and its validity assessed.

During the initial phase of method development and validation it became clear that microplastic

analysis is not nearly as straight-forward as much of the literature leads to believe. This obser-

vation is in line with what a few other researchers have reported during the latter years, for

example Löder and Gerdts (2015), who showed that many of the particles which by the light-

microscope assisted eye might be identified as microplastic in reality are mineral particles such

as quartz with a high diffraction index. A similar conclusion was made in the present study,

clearly ruling out light microscopy as an analytical method for determining microplastic particles

<500 µm in environmental samples. Hence, in the present study light microscopy was not a

valid method for identification of particles of unknown material, size, shape, and origin. Light

microscopy is, though, applicable when studying systems where distinct plastic particles, for

example strongly colored or fluorescence particles, are spiked and then recovered.

In general, there is consensus that the most appropriate and effective method for identifying

both size and material of microplastic particles is the use of Fourier Transform – Infrared (FT-

IR) spectroscopy, preferable as an imaging system where a FT-IR spectroscope is combined

with a microscope equipped with, for example, a Focal Plane Array (FPA), hereby allowing what

typically is called FT-IR imaging (Vianello et al., 2013; Loder et al., 2015; Tagg et al., 2015).

However, the number of such studies is very small, as the equipment is expensive and requires

highly trained personnel (Rocha-Santos and Duarte, 2015). Consequently there is no standard-

ized and generally accepted method on how an environmental sample is to be analysed for

microplastic.

Another issue that became obvious through the study is that a FT-IR spectrum obtained for a

particle should not simply be compared to a standard spectrum from a material database for

polymer identification. While such comparison does assist the analysis, it leaves room for de-

tection errors, i.e. there is a high risk of obtaining false positive or false negative particle materi-

al identifications. Instead, a spectral analysis has to be performed for each particle in a sample,

to identify which chemical bonds are present and from this information conclude what material it

is made off. Ignoring this latter step increases the risk of not detecting for example co-polymers

and particles with deteriorated spectra. It also increases the risk for misinterpreting the material

of the particle, for example so that natural organic particles are identified as plastic particles or

that the type of plastic is misinterpreted. In this context it was our experience, that spectral

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

analysis requires an understanding of the chemistry of polymers and personnel thoroughly

trained in infrared spectral analysis.

Before analysing a sample on a FT-IR imaging system, irrelevant sample constituents must be

removed and the microplastic concentrated. There are a limited number of studies that have

addressed this in terms of making a sample suitable for FT-IR analysis, for example to extract

microplastic from beach sand (Cauwenberghe et al., 2015) and also some studies on the prep-

aration of wastewater samples (Mintenig et al., 2014). However, there is no clear consensus

and no clearly defined method on how this should be done.

When it comes to assessing the uncertainty of a microplastic determination, similar to what is

known from analytical chemistry, the knowledge is still more erratic. Most studies have simply

ignored the issue of sampling and measurement uncertainty, and none have presented a sys-

tematic way to assess such uncertainty.

3.2

The methodology at a glance

The current study addresses microplastic particles in the size range 20-500 μm and all steps of

the methodology are optimized towards this range. The method applied for detection of micro-

plastic in an environmental sample is divided into 5 major steps:

1. Sample collection

This step attempts to collect samples which are representative of the environmental

system analyzed. In this study, 4 types of samples are collected: Raw wastewater,

treated wastewater, wastewater sludge, and agricultural soils.

Sample preparation and concentration

Ideally speaking, the purpose of this step is to remove all non-plastic particles and all

other substances that might interfere with the following analytical steps. The method-

ology for the sample preparation differs between sample types and the sample prepa-

ration for wastewater, sludge and soil are quite different. The preparations do though

all end up in similar products, namely a concentrate of particles that can be analysed

on a FT-IR imaging system.

Sample analysis applying FT-IR imaging

A sub-sample of the particle concentrate is transferred to a FT-IR imaging system and

analysed: The sub-sample is placed on a microscope slide where it is illuminated by IR

light. The spectrum of the transmitted (or reflected) IR light is analysed and a spectrum

created, which is characteristic for the material of the investigated particle. The FT-IR

imaging system scans the slide with a resolution of some micrometres and produces

hereby a FT-IR spectral map of the scanned area. At the same time a traditional light

microscope image is produced, which allows visual inspection of the same area of the

slide as well as determination of the particle size. Figure 4 illustrates this principle

where the upper picture is a visual map of the FT-IR scanned area shown in the pic-

ture below.

Interpretation of the infrared spectra applying spectral analysis techniques

The map of FT-IR spectra is manually processed by spectral analysis. First a rough

analysis is done, rejecting all particles that are of materials that cannot be plastic (for

example inorganic particles). This leads to rejection of the majority of the particles as

sample preparation is not able to remove all particles of natural origin. This is illustrat-

ed by the highlighted blue area of the spectrum shown in Figure 4. All plastic polymers

must absorb IR light in the blue spectral range, and particles that do not absorb at this

wavenumber are hence rejected up front. The possible plastic candidates then under-

go manual spectral analysis to identify the nature of their chemical bonds and here

through their composition.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Analysing results and calculating concentrations

At the same time as the material of a particle was determined, its size was noted

down. From this its volume was calculated and multiplied by the density of the plastic

material, leading to the mass of the particle. Concentrations were then calculated by

relating the area scanned on the FT-IR imaging system to the total volume taken into

analysis.

Figure 4. Searching for plastic particles on a 700x2800 μm section of an IR transmission

window. The upper picture is a light microscopy image, the middle picture an IR heat

map. The cross-mark of the two upper pictures shows the particle for which the spec-

trum in the lower picture is created.

Sampling volumes

For the raw wastewater, 1 L of sample was taken into work and pre-filtered over a 500 μm

sieve. A 200 mL subsample of the filtrate was treated and the microplastic concentrated in 5 mL

ethanol. For the treated wastewater, the water was filtered on site over 3 steel filters of 10 μm

mesh size until these clogged. The amount of treated wastewater that could pass the filters

before they clogged was between 4.1 and 81.5 L. The 3 filters where then treated and the mi-

croplastic concentrated in 5 mL of ethanol. For sludge, approx. 1 kg was collected and a sub-

sample of 0.1 g was taken here from. The subsample was treated and the microplastic concen-

trated in 5 mL ethanol. For soil, cores of approx. 300 mL were collected, a sub-sample of 50 g

treated and the microplastic concentrated in 5 mL ethanol. Depending on the type of sample,

between 0.02 and 0.3 mL of the ethanol particle suspensions were transferred to the FT-IR

imaging system for analysis.

Further details on sampling, sample preparation, FT-IR imaging, and spectral analysis are given

in Appendix 1.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

3.3

Method validation

Due to the nature of the targeted pollutants, methods can only be validated to a certain extent

and uncertainties can only be estimated. There are several reasons for this:



Dissimilar to analysis of a dissolved chemical compound in an environmental matrix,

the targeted pollutant is not a well-defined substance. Plastic is a wide range of

manmade polymers and also some natural polymers such as natural rubber. The

boundary between what is plastic and what is not is a question of definition e.g. in con-

connection with paint particles, composite materials, semi-synthetic fibers (such as

viscose-rayon) and composite materials between such products and more traditional

plastic polymers. In the present method we follow the definition by Lassen et al.

(2015), namely plastics are solid materials formed from polymers of a mainly petro-

chemical origin, a definition that includes rubbers and paints.



The issue that plastic is not a well-defined substance further leads to issues when ap-

plying standard analytical validation techniques, like spiking of a sample, to validate

the method. What shall one spike with in order to validate methods?



In addition to being made of a multitude of materials, plastic particles come in all

shapes and sizes. This creates significant problems when validating analytical meth-

ods. So will, for example, the recovery of a plastic particle depend on all three parame-

ters: material, size, and shape. In addition hereto, the sample preparation and concen-

tration techniques will affect the plastic particles differently, depending again on mate-

rial, size, and shape.

In a similar way, detection limits are problematic to define when identifying particles of a wide

range of shapes, sizes and materials. Nevertheless, it is highly important to attempt to quantify

the validity of the microplastic determination, its uncertainties, and its detection limit.

In the present study, the extraction method was validated by adding a known number of micro-

plastic particles to raw wastewater and counting the recovered particles after sample prepara-

tion. The materials used were spherical polystyrene (PS) beads of 100 μm diameter, and high

density polyethylene (HDPE) particles and styrene butadiene rubber (SBR) particles of 80-150

μm. The latter two were made by grinding down larger plastic pieces and sieving the material

into appropriate sizes. The results showed recovery rates slightly below 100% for PS (78%

±17%) and HDPE (61% ±29%). For SBR it was slightly above 100% (120% ±38%) (Figure 5).

The latter was deemed due to the SBR particles easily disintegrating into smaller particles. The

difference between the recoveries was though not statistically significant.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 5. Recovery of microplastic particles added to raw wastewater

Detection limits

The detection limit of the method depends on the fraction of the sample scanned by the FT-IR

imaging system. In the present study, we have scanned till at least 10 plastic particles had been

positively identified, or till we had scanned an area of at least 4 mm out of the 78 mm on the

microscope slide.



For wastewater samples the detection limit was better than 4 μg/L.



For treated wastewater the detection limit was better than 0.20 μg/L.



For sludge the detection limit was better than 20 μg/g.



For soil the detection limit was better than 0.04 μg/g.

Details on the method validation, detection limits and uncertainties are presented in Appendix 2.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

4. Results and discussion

4.1

Microplastic in Danish wastewater treatment plants

This study addresses 10 wastewater treatment plants where microplastic in the size range 20-

500 μm was analysed in the inlet and outlet wastewater. In addition, anaerobic digested sludge

from 5 of these plants was analysed for microplastic in the same size range. The samples were

all analysed for microplastic occurrence and which polymers the microplastic particles were

made from.

4.1.1

Microplastic concentrations in raw and treated wastewater

The mass of microplastic in the raw and treated wastewater is shown in Table 2 while the cor-

responding numbers are shown in Table 3. Figure 6 and Figure 7 illustrate the numbers graph-

ically. For both types of sample there is a quite large difference between the average and the

median, indicating that the dataset was not normal distributed. In more concrete terms the dis-

crepancy was caused by one treatment plant having much higher concentrations of microplastic

than all the other WWTPs. It seems likely that this was caused by random variation in the sam-

pling and the following analysis. Hence the median should be used and not the average. This

issue is not observed for the numbers of particles.

The variability between the 10 treatment plants was quite large. The plant with the least micro-

plastic in the raw wastewater had 0.22 mg/L while the one with the largest mass had 29.6 mg/L.

The variability in statistical terms, i.e. 25 and 75 percentiles, is shown in Table 2. Corre-

spondingly, the variability in number of particles was from 13,000 to 442,000 particles per litre,

with the statistic variability shown in Table 3. This variability is probably due to real differences

between the wastewater entering the treatment plants in combination with analytical uncertain-

ty. A similar variability was seen for the treated wastewater. Such large variability is not un-

common for organic micropollutants where concentration ranges of several decades are often

seen (Luo et al. 2014). However, the target of the present study was not to assess individual

treatment plants but to assess the general median of Danish wastewater.

Table 2. Plastic mass in raw and treated wastewater. Average and median of 10 treatment

plants as well as the 25 and 75 percentile of the dataset.

Average

Raw wastewater [mg/L]

Treated wastewater [mg/L]

8.0

0.034

Median

5.9

0.016

percentile

2.2

0.0047

percentile

0.037

Table 3. Plastic particle numbers in raw and treated wastewater. Average and median of

10 treatment plants as well as the 25 and 75 percentile of the dataset.

Average

Raw wastewater [no./L]

Treated wastewater [no./L]

127,000

5,800

Median

86,000

6,400

percentile

70,000

4,400

percentile

130,000

8,000

Typical wastewater holds 320-740 mg COD/L and 190-450 mg SS/L. The amount of micro-

plastic in the size range 20-500 μm hence accounts for roughly 1% of the total organic matter of

the raw wastewater and 2-3% of the total amount of suspended solids.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 6. Microplastic mass concentrations in raw wastewater versus treated

wastewater. Note the logarithmic scale on the y-axis.

Figure 7. Microplastic particle number concentrations in raw wastewater versus treated

wastewater. Note the logarithmic scale on the y-axis.

Table 2 and Table 3 tell that the treatment plants released 0.3% of the mass of microplastic

they received and 7.4% of the particle numbers. The variability around these numbers is though

high and the results should hence only be seen as indicative. The difference in efficiency be-

tween microplastics measured as numbers compared to microplastic measured as mass might

be due to the primary settling tanks of the treatment plants having a larger effect towards larger

particles compared to smaller particles. Another reason might be that larger plastic particles to

some degree are broken down while in the treatment plant. Nevertheless, for plastic measured

in terms of both mass and numbers, the overall efficiency to remove microplastic particles was

high compared to most dissolved substances that occur in municipal wastewater.

4.1.2

Size distributions of microplastic particles

The median size of microplastic particles in the treated wastewater was approx. 20% smaller

than in the raw wastewater. Figure 8 and Figure 9 show the size distributions of particles in the

raw and treated wastewater, respectively. The figures show the largest and smallest measured

diameter of each plastic particle as well as the average of the two diameters. The median of

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

plastic particle size in the raw wastewater was 50 μm while it was 41.5 μm in the treated

wastewater. The corresponding values for the 75-percentiles (i.e. the 75% of the smallest parti-

cles) were 65 μm in the raw wastewater and 55.9 μm in the treated wastewater. Looking at the

shape of individual particles, the ratio between largest and smallest diameter of a particle dif-

fered slightly between the raw and treated wastewater, namely respectively 1.22 and 1.41.

Figure 8. Size distribution of microplastic particles in the raw wastewater. Three values

are given: The smallest measured diameter of a particle, the average of its two measured

diameters, and its largest measured diameter.

Figure 9. Size distribution of microplastic particles in the treated wastewater. Three val-

ues are given: The smallest measured diameter of a particle, the average of its two

measured diameters, and its largest measured diameter.

Even though the difference in particle size was not substantial, the difference in mass is signifi-

cantly more so because the volume of particles comes in the third power of their size. Figure 10

compares the particle masses for the raw and treated wastewater. The median particle mass in

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

the raw wastewater was twice that of the treated wastewater (41 ng versus 20 ng). For particle

masses less than approx. 7 ng there seems though to have been slightly more particles in the

raw wastewater compared to the treated. The reason is unknown, but could relate to technical

aspects of the treatment plants or to breakdown of particles that affect smaller particles more

than larger ones. It could, though, also be a random artefact of analytical uncertainties.

Figure 10. Mass distribution of microplastic particles in the raw and treated wastewater

The raw wastewater furthermore held some rather large particles that accounted for a signifi-

cant fraction of the total microplastic mass. The 4 largest particles (Figure 10) account for 35%

of all the plastic mass found in the raw wastewater samples.

The observations made in this study lead to the conclusion that when addressing the efficiency

of wastewater treatment plants to retain microplastic, mass as the unit of measurement is signif-

icantly more reproducible than particle numbers. The number of particles is affected by physical

breakdown processes, and this breakdown can result in increases in particle numbers without

an increase in plastic mass. Hence, when applying only particle numbers for quantifying the

efficiency of a treatment system, this system could in principle ‘produce’ microplastic because

larger particles were broken down into smaller particles. On the other hand, when it comes to

the impact of microplastic on aquatic fauna, the number of particles potentially plays a signifi-

cant role. Hence microplastic mass should be used to assess treatment efficiencies and particle

numbers should additionally be reported to support environmental impact assessment.

4.1.3

Microplastic concentrations in sludge

The mass and particle numbers found in digested sludge from 5 treatment plants are shown in

Table 4 and Table 5. Digested wastewater sludge typically has a dry matter content of 25-30%,

and the median of the measurements hence indicate that approx. 2% of the total dry matter

content of the sludge was microplastic.

Table 4. Plastic mass in digested wastewater sludge. Average and median of 5 treatment

plants as well as the 25 and 75 percentile of the dataset.

Average

Sludge [mg/g]

4.5

Median

6.5

percentile

2.0

percentile

6.5

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Table 5. Plastic particle numbers in digested wastewater sludge. Average and median of

5 treatment plants as well as the 25 and 75 percentile of the dataset.

Average

Sludge [no./g]

169,000

Median

158,000

percentile

79,000

percentile

175,000

4.1.4

Mass balance of microplastic in wastewater treatment plants

A rough mass balance can be made on the inlet and outlet mass of microplastic. Assuming that

microplastic is inert in the treatment plant, the total mass in the inlet must equal the sum of the

mass in the sludge and in the discharged wastewater.

The total amount of wastewater entering the Danish treatment plants in the period 2012-2014

was 686 million m per year (Naturstyrelsen, 2013; 2015a; 2015b). Assuming that the median

concentration of dry weather microplastic concentrations in the raw wastewater are representa-

tive for the annual load on treatment plants, the corresponding microplastic load on the plants

was 4,035 ton/year (25 and 75 percentiles: 1124 and 5072 ton/year). Hereof 11 ton/year (25

and 75 percentiles: 4.9 and 16 ton/year) was discharged with the treated wastewater to the

receiving waters, leaving 4,024 ton/year not emitted to the aquatic environment. These num-

bers are subject to some uncertainty. One uncertainty relates to the samples representing dry-

weather wastewater only and that concentrations during storm runoff most likely are different

from those at dry weather. Another uncertainty is the analytical uncertainty of the study. The

size of both these uncertainties is basically unknown.

The total amount of sludge produced in Denmark is approx. 132,600 ton-DM/year (Miljøstyrel-

sen, 2009). Digested sludge typically has a dry matter content of 25-30%, and the correspond-

ing amount of dewatered sludge is hence approx. 480,000 ton-dewatered-sludge/year. Here of

a median of 0.7% is microplastic. Applying the median value of the microplastic mass found in

this study, this corresponds to a microplastic mass in the sludge of approx. 3,100 ton/year with

25 and 75 percentiles of 970 and 3,110 ton/year. Also this number is subject to uncertainty of

an unknown magnitude. However, the mass of microplastic found in the sludge amounted to

approx. ¾ of the microplastic in the inlet, which gives confidence in the number as the mass in

the sludge must be equal to or less than the mass in the inlet.

Adding up the mass balance showed that approx. ¾ of the total microplastic load on the treat-

ment plants could be accounted for. It is not known whether the lacking ¼ of the microplastic

mass is simply due to measurement uncertainties, or if it is caused by degradation of the nylon

particles in the digesters, as the proportion of nylon decreased significantly in the sludge-

fraction (see further discussion in 4.1.6).

Lassen et al. (2015) estimated that some 2,000-5,600 ton/year of microplastic (size range 1µm -

5 mm) was discharged to sewerage and hence end up at wastewater treatment plants. Taking

all the uncertainties of a literature-based assessment of microplastic loads into account, this

finding is in good agreement with the values measured in the present screening study. There

are though some significant discrepancies in what was predicted (not measured) by Lassen et

al. (2015) in terms of polymer materials. Lassen et al. predicted that 1,600-2,500 ton/year of tire

particles (styrene butadiene co-polymers) should be discharged to wastewater treatment plants.

A likely explanation for these not being found in the wastewater is that they were below the size

limit of the present study (20 μm). So did for example both Dall’Osto et al. (2014) and

Mathissen et al. (2011) in their studies report the majority of particles from car tire abrasion

between 10 and 100 nm. Nevertheless, the issue about the fate of particles created by care tire

abrasion does lead to open and unsolved questions which should be addressed by future stud-

ies.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

4.1.5

Microplastic loads on the aquatic environment

The literature study conducted by Lassen et al. (2015) had estimated that the total Danish

emission of microplastic (size range 1µm - 5mm) to the aquatic environment is in the range

600-3,100 ton/year. Comparing this to the findings of the present study of approx. 11 ton/year

(25 and 75 percentiles: 4.9 and 16 ton/year) (size range 20-500 μm) discharged with the

treated wastewater, it can be concluded that treated wastewater from Danish municipal treat-

ment plants constitute a minor fraction of the total amount of microplastic released to the aquat-

ic environment.

4.1.6

The polymer composition of the microplastic particles

The by far most common plastic material in the raw wastewater was types of polyamide/nylon

(Fejl!

Henvisningskilde ikke fundet.

and Figure 12). It seems reasonable to assume that a

probable major origin for these plastics is various forms of textiles, clothing and carpets. Other

types of plastic belong to the group of polyethylene and co-polymers (PE), polypropylene and

co-polymers (PP), and PVC. PE is used in for example packaging like plastic bags, plastic films,

plastic bottles, and so on. PP is also used in packaging as well as in for example textiles includ-

ing cloth and carpets. PVC is more rigid than PE and PP and used in for example construction

materials, non-food packing, and electrical cable insulations. In addition to PE, PP, and PVC,

there was in a few samples found a high abundance of zinc stearate coated particles. Zinc

stearate is strongly hydrophobic and used to coat a range of plastics to enhance the product’s

behaviour. When a plastic particle is coated with zinc stearate it is not possible to identify the

underlying plastic material but zinc stearate coating can occur on most popular household ap-

pliance plastics such as PP, PE, and PS (polystyrene).

The total distribution of plastic polymers in raw wastewater with respect to the identified mass is

shown in

Fejl! Henvisningskilde ikke fundet.

while the distribution with respect to the number

of particles found of each material is shown in Figure 12. Polyamide/nylon was dominant both in

terms of particle numbers and particle mass. Comparing the two figures indicates that the rela-

tive distribution of polymers on particle mass and number differed somewhat for polymers other

than polyamide/nylon. Especially for zinc stearate coated particles indicating that these were

mainly small particles. However, the number of particles behind those fractions is comparatively

small and it cannot be excluded that this variation is due to random variability.

Figure 11. Distribution of the mass of plastic polymers in raw wastewater. A total of 181

plastic particles were identified with respect to their material.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 12. Distribution of the particle numbers on identified plastic polymers in raw

wastewater. A total of 181 plastic particles were identified with respect to their material.

Figure 13 and Figure 14 show the similar data for the treated wastewater. Also here, it was

polyamide/nylon that dominated the picture in terms of polymers, followed by PE material types.

Furthermore, comparing the composition between inlet and outlet in the investigated size range

of microplastic particles it is seen that the treatment plant does not to any significant extend

preferentially remove specific plastic polymers, in other words, the removal efficiency for all

polymers are more or less the same. However in the treated samples PVC was not found.

Whether this is due to the circumstance that a limited number of outlet samples has been ana-

lysed, or due to preferential removal of particles, is unknown.

Figure 13. Distribution of the mass of plastic polymers in treated wastewater. A total of

150 plastic particles were identified with respect to their material.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 14. Distribution of the particle numbers on identified plastic polymers in treated

wastewater. A total of 150 plastic particles were identified with respect to their material.

Figure 15 and Figure 16 show the distribution of plastic polymers found in the digested

wastewater sludge. The total number of microplastic particles identified was comparatively low

(29). This was partly due to fewer samples being analysed compared to raw and treated

wastewater and partly due to time limitations of the study. Of the 29 particles, PE was the dom-

inating material followed by nylon which was significantly lower in distribution than observed for

the wastewater samples. This could indicate that the anaerobic digestion process affects the

plastic, either by breaking it down to particles too small to detect by the applied approach, or by

biological degradation of nylon, which decreases significantly in distribution of the plastic parti-

cles in the sludge-fraction as compared to the wastewater. Bacterial digestion of nylon has

been shown in environments with sufficient nylon by bacteria producing a nylon oligomer hydro-

lase (Gautam et al, 2007). Whether or not this is the cause of the discrepancy between the

mass distributions in the raw wastewater and the digested sludge is not known and needs fur-

ther investigation. Due to a limited sample size, it cannot be excluded that the observed differ-

ences are due to random variability.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 15. Distribution of the mass of plastic polymers in digested wastewater sludge. A

total of 29 plastic particles were identified with respect to their material.

Figure 16. Distribution of the particle numbers on identified plastic polymers in digested

wastewater sludge. A total of 29 plastic particles were identified with respect to their

material.

It is noteworthy that the study did not reveal a single rubber particle from tire abrasion even

though a literature based study on microplastic releases in Denmark had indicated tire abrasion

as the largest microplastic source released to the Danish environment (Lassen et al., 2015).

The most likely reason is that such particles were smaller than the 20 μm, which were the lower

size limit of the present study. Mintenig et al. (2016) applied a comparable FT-IR technique on

treated wastewater from German WWTPs. Comparing the identified plastic polymers to those

found in that study, some significant differences were seen. For particles <500 μm, they found

polyethylene (PE) to dominate the samples (40% of particle numbers), followed by polyvinyl

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

alcohol (16%) and then polyamide (PA, what we have termed nylon or polyamide/nylon in this

report) and polystyrene (PS, 8% of particle numbers each). However, similarly to the present

study, Mintenig et al. (2016) did not find rubber (styrene-butadiene co-polymers) in any of their

samples – neither in the smaller fraction (20-500 µm), which they measured by an FT-IR imag-

ing technique similar to the present study, or in the larger size fractions (>500 µm), which they

measured by FT-IR-ATR similar to what was done in the present study to validate spectra from

car tires (Appendix 2.4).

4.2

Occurrences and sources of microplastic on Danish

farmlands

Microplastic was measured in 5 samples from farmlands that had received sludge as fertilizer

and 5 farmlands that had not. The latter soils have furthermore never received any other organ-

ic waste product of non-farming origin. The concentration of microplastic in these samples is

shown in Table 6 and Table 7. The number of plastic particles in both types of soils was low

and the soils were rather comparable with respect to the microplastic concentrations found. The

soils that had not received sludge had twice the microplastic content than the soils that had not.

This difference is most likely due to uncertainties in sampling and measurement as the total

number of detected particles was rather low (13 and 24 particles in the soil with and without

sludge, respectively).

Table 6. Plastic mass in farmland soils. Average and median of 25 soils as well as the 5

and 75 percentile of the dataset.

Average

Soils with sludge [mg/kg]

Soils without sludge [mg/kg]

6.2

Median

5.8

percentile

1.4

4. 4

percentile

7.6

Table 7. Plastic particle numbers in farmland soils. Average and median of 5 soils as well

as the 25 and 75 percentile of the dataset.

Average

Soils with sludge [no./kg]

Soils without sludge [no./kg]

82,000

236,000

Median

71,000

145,000

percentile

29,000

53,000

percentile

143,000

436,000

The concentration of microplastic in the soils are deemed low, namely around 10 mg/kg. Such

concentration is comparable to, for example, the background concentrations of heavy metals in

Danish soils (lead, copper, zinc, chromium, nickel, or cobalt found; By- og Landskabsstyrelsen

(2010)).

The composition of the plastic material found in the soils is shown in Figure 17 and Figure 18.

The dominant polymers were PE and PP. Both of these polymers can origin from a wide pallet

of materials, including packaging materials applied for agricultural purposes. The soils that had

received sludge had both of these polymers in significant concentrations while the samples that

had not received sludge only had a low concentration of PP. Nylons was present in relatively

low concentration in both samples as compared with the wastewater and sludge samples. The

high proportions of PP in soils with sludge may indicate that PP has a higher persistence to

disruption and degradation than the other plastic polymers; this however needs further investi-

gations and a larger sample size to further evaluate.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 17. Distribution of the mass of plastic polymers in soils that had received sludge.

A total of 13 plastic particles were identified with respect to their material.

Figure 18. Distribution of the mass of plastic polymers in soils that had not received

sludge. A total of 24 plastic particles were identified with respect to their material.

4.2.1

The impact of wastewater sludge on agricultural soil

The amount of wastewater sludge that can be applied on Danish agricultural soil corresponds to

90 kg of phosphorous per hectare calculated as an average over 3 years. This typically means

that farmland will receive one load of sludge every 3 years. Assuming average values of dry

matter content and phosphorous content as reported by Miljøstyrelsen (2013), these 90 kg of

phosphorous per hectare correspond to approx. 10.4 ton of dewatered sludge per hectare.

Assuming the microplastic content found in the present study (Table 4) and assuming a tilling

depth of 30 cm (i.e. the soil depth into which the wastewater sludge is mixed) leads to the one

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

batch of sludge applied every 3 year to increase the microplastic concentration of the soil by

approx. 15 mg/kg (25 and 75 percentiles: 6.7 and 22 mg/kg). In other words, one load of

sludge contributes by approximately the same amount as found as background concentration in

the soils that had not received sludge (Table 7). This estimate is rather rough and not a true

mass balance, and there are significant uncertainties in both the estimate on microplastic occur-

rence in soils that have not received sludge and on microplastic in sludge. Nevertheless, the

estimate indicates that while sludge application does contribute to the microplastic content of

farm lands, it is not the only source of significance to microplastic in agricultural soils.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

5. Perspectives

The results from the present screening study show that the emission of microplastic from Dan-

ish WWTPs to the aquatic environment is minor compared to the total load on the treatment

plants. It represents at worst case approximately 3% of the total mass of microplastic released

to the Danish aquatic environment. However, it has to be stressed that the results from this

study is only indicative as the investigation has been a screening study and only addressing the

size range 20-500 μm. Deeper sampling and analysis at each WWTP with coverage of the full

range from 1 to 5000 μm is necessary to get a thorough understanding of the variation between

WWTPs and to obtain statistically conclusive results.

Tertiary filter technologies are available on the market to further minimize the microplastic

emission to the aquatic environment from the WWTPs. However, due to the already high re-

moval efficiency of the wastewater treatment technologies routinely implemented in Denmark,

other aspects of the urban sewerage system are likely more important with respect to discharge

of microplastic into the aquatic environment. For example microplastic particles discharged via

combined sewer overflows, via misconnected wastewater or via stormwater discharges.

The future focus on microplastic in wastewater is recommended to be directed to the sludge

fraction as almost all microplastic mass entering the Danish wastewater treatment plants ends

up in the sludge (>99%) and thus the farmlands where sludge is used as fertilizer. This study

further indicates that the impact of microplastic in sludge on sludge-fertilized farmlands is low

and that other sources might be of significance for the microplastic found in farmlands. There

are, though, a number of open questions in this respect and future investigations are needed to

further investigate other sources of microplastic in Danish soils and whether or not the plastic

particles accumulate in the soil over time.

The sources of the detected plastic polymers are difficult to identify from the present study, but

future investigations should look into the possible origins of polyamide/nylon, polyethylene and

polypropylene, which were the most dominant polymers detected in the samples. The most

abundant polymer in the wastewater was polyamide/nylon, which most probably originates from

textiles, clothing and carpets, while polyethylene and polypropylene probably originates from

different types of packaging or similar. The lack of detected rubber particles from tires must be

further investigated. While it has never been identified in any study addressing microplastic

polymers in wastewater samples, it has in mass balance based literature studies been identified

as an abundant microplastic pollutant. Plastic from personal care products was not detected,

which verifies previous reports that it contributes insignificantly to the total emission of micro-

plastics in Denmark (0.2%, Lassen et al., 2015).

Some plastic polymers may be biologically degradable under the right conditions e.g. nylon and

polyurethane. This agitates for development of special adapted treatment technologies of

wastewater sludge, where the optimal conditions for biological or thermal degradation ensure a

minimal emission of microplastic to the terrestrial environment. Further investigations should

moreover evaluate the degree of plastic degradation in the terrestrial environment.

With the experience from the method development and optimisation of a valid approach for

microplastic identification and quantification carried out in the present study, it is evident that

microplastic results reported in literature should be carefully reviewed and the method used for

detection of microplastic should be looked over before acknowledging the reported results. We

have learned that light microscopy alone is unsuitable for investigations of microplastic in envi-

ronmental samples and verification of the particle material as plastic is highly important. There-

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

fore, we recommend that future investigations of microplastic in environmental samples should

be realised with either FT-IR spectroscopy methods and/or possibly Raman spectroscopy

methods. We further recommend that sampling methods and analytical methods for micro-

plastic analysis is standardized to allow comparison between results of microplastic monitoring

and investigations.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Acknowledgement

This investigation has only been possible due to the very friendly and collaborative personnel at

the 10 investigated WWTPs, Damhusåen, Ejby Mølle, Fredericia, Herning, Horsens, Kolding,

Lynetten, Marselisborg, Vejle and Aalborg West. We deeply thank their understanding and

patience in connection with the sample collection.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

References

By- og Landsskabsstyrelsen (2010). Kemiske stoffer – Vurdering af stoffer i forhold til farlighed i

grundvandet. By- og landskabstyrelsen, Miljøministeriet, ISBN: 978-87-92617-87-3

Dall’Osto M, Beddows DCS, Gietl JK,. Olatunbosun OA, Yang X, Harrison RM (2014). Charac-

teristics of tyre dust in polluted air: Studies by single particle mass spectrometry (ATOFMS).

Atmospheric Environment 94_ 224-230

Gautam R, Bassi AS, Yanful EK (2007). A review of biodegradation of synthetic plastic and

foams. Applied Biochemistry and Biotechnology, 141(1): 85–108

Lassen C., Hansen S.F., Magnusson K., Norén F.,. Hartmann N.I.B, Jensen P.R., Nielsen T.G.

and A. Brinch (2015) Microplastics – Occurrence, effects and sources of releases to the en-

vironment in Denmark. The Danish Environmental Protection Agency. Environmental project

No. 1793.

Löder MGJ, Kuczera M, Mintenig S, Lorenz C, Gerdts G (2015). Focal plane array detector-

based micro-Fourier-transform infrared imaging for the analysis of microplastics in environ-

mental samples. Environmental Chemistry, 12(5): 563-581

Löder MGJ, Gerdts G (2015).Methodology Used for the Detection and Identification of Micro-

plastics—A Critical Appraisal. Chapter 8 of Marine Anthropogenic Litter, Editors: Bergmann

M, Gutow L, Klages M. ISBN 978-3-319-16509-7

Luo Y, Guo W, Ngo HH, Nghiem LD, Hai FI, Zhang J, Liang S, Wang XC (2014). A review on

the occurrence of micropollutants in the aquatic environment and their fate and removal dur-

ing wastewater treatment. Science of the Total Environment, 473–474: 619–641

Magnusson K. and C. Wahlberg (2014) Mikroskopiska skräppartiklar i vatten från avloppsren-

ingsverk. IVL Svenska Miljöinstitutet.

Mathissen M, Scheer V, Vogt R, Benter T (2011). Investigation on the potential generation of

ultrafine particles. Atmospheric Environment 45: 6172-6179

from the tireeroad interface

Miljøstyrelsen (2009). Spildevandsslam fra kommunale og private renseanlæg i 2005. Oriente-

ring fra Miljøstyrelsen Nr. 3

Miljøstyrelsen (2013). Livscyklusvurdering og samfundsøkonomisk analyse for anvendelse af

spildevandsslam. Miljøprojekt nr. 1459, 2013

Mintenig S, Int-Veen I, Löder M, Gerdts G (2014). Mikroplastik in ausgewählten Kläranlagen

des Oldenburgisch- Ostfriesischen Wasserverbandes (OOWV) in Niedersachsen. Pro-

benanalyse mittels Mikro-FT-IR Spektroskopie. Abschlussbericht. Report for Oldenburgisch-

Ostfriesischer Wasserverband (OOWV) and Niedersächsischer Landesbetrieb für Wasser-

wirtschaft, Küsten- und Naturschutz (NLWKN)

Mintenig SM, Int-Veen I, Löder MGJ, Primpke S, Gerdts G (2016). Identification of microplastic

in effluents of waste water treatment plants using focal plane array-based micro-Fourier-

transform infrared imaging. Water Research, doi: 10.1016/j.watres.2016.11.015.

Murakami S. (2009). Points to Note in Rubber Analysis: Black Rubber. Tokyo: Shimadzu. Re-

trieved from http://www.shimadzu.com/an/FT-IR/support/tips/letter11/rubber.html

Naturstyrelsen (2013). Punktkilder 2012.

Naturstyrelsen (2015a). Punktkilder 2013.

Naturstyrelsen (2015b). Punktkilder 2014.

Pedersen B.M. and M. Winther-Nielsen (2015), Mikroplast i spildevand fra renseanlæg, littera-

turindsamling. DANVA rapport Maj 2015.

Rocha-Santos T, Duarte AC (2015). A critical overview of the analytical approaches to the oc-

currence, the fate and the behavior of microplastics in the environment. TrAC Trends in

Analytical Chemistry, 65: 47–53

Sckerl S.Aa. (2012) Historisk stor andel af spildevandsslam til jordbrug, Teknik & Miljø, Februar

2012 p. 54-55.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Tagg AS, Sapp M, Harrison JP, Ojeda JJ (2015). Identification and Quantification of Microplas-

tics in Wastewater Using Focal Plane Array-Based Reflectance Micro-FT-IR Imaging. Ana-

lytical Chemistry, 87(12): 6032-6040

Thompson RC et al (2004) Science, 304, p. 838-.

Van Cauwenberghe L, Devriese L, Galgani F, Robbens J, Janssen CR (2015).Microplastics in

sediments: A review of techniques, occurrence and effects. Marine Environmental Re-

search, 111: 5-17 SI

Vianello A, Boldrin A. Guerriero P, Moschino V, Rella R. Sturaro A, Da Ros L (2013). Micro-

plastic particles in sediments of Lagoon of Venice, Italy: First observations on occurrence,

spatial patterns and identification. Estuarine, Coastal and Shelf Science, 130: 54-61

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Appendix 1.

Analytical

method

Appendix 1.1

Sampling

The risk of contamination with plastic in the sampling procedure was minimized by using glass

and metal equipment. When equipment with plastic could not be avoided the contamination risk

was evaluated and when possible the contaminating plastic was subtracted from the final re-

sults.

Raw wastewater

Raw wastewater was collected at three independent sampling events at each of the ten

wastewater treatment plants (Table 1) giving a total of thirty raw wastewater samples. The

samples were flow proportional 24 h samples collected in the inlet wastewater stream with the

auto samplers of the treatment plants. The samples were collected at dry weather and the max-

imum of rain allowed was 3 mm 48 h before and during the sampling, 72 h in total.

The auto samplers have plastic tubing and the wastewater was collected in big plastic bottles

and after 24 h, 1 litre of the collected wastewater was transferred and stored in a glass jar. The

amount of contamination with microplastic during the sampling procedure was assumed to be

minimal compared to the microplastic present in the raw wastewater.

Treated wastewater

Treated wastewater was collected at three independent sampling events at each of the ten

wastewater treatment plants (Table 1) giving a total of thirty treated wastewater samples. The

samples were collected by filtration of outlet water through 10 µm stainless steel filters until the

filter clogged. To ensure enough particles for investigation three filters were clogged at each

sampling event. The volume of treated wastewater necessary to filter differed from plant to plant

(0.5-108 litres per filter). The samples were collected at dry weather and the maximum of rain

allowed was 3 mm 48 h before sampling.

All equipment used to collect treated wastewater was of glass and metal and no risk of contam-

ination with microplastic from the sampling procedure is expected.

Sludge

Sludge samples were collected at the final dewatering unit after the sludge digestion at the ten

investigated wastewater treatment plants (Table 1). The sludge was collected at two independ-

ent sampling events giving a total of twenty sludge samples of about 1 kilogram. The period

between the two sampling events were longer than the sludge age at the specific treatment

plants plus the sludge retention time in the digester, thereby ensuring that the sludge samples

were indeed true biological replicates.

All equipment used to collect the sludge samples was of glass and metal and no risk of contam-

ination with microplastic from the sampling procedure is expected.

Soil from farmlands

Soil samples were collected in the early spring shortly after the frost had left the soil.

Four soil samples were taken from each of 5 fields that had received sewage sludge and an-

other 4 soil samples were taken from each of 5 other fields that had never received sewage

sludge as fertilizer. The 2x5 fields were all geographically close to each other in Northern Jut-

land. The 4 soil samples from each field were taken approx. 100 m apart and at least 20 m into

the field. The samples were collected as columns with a planting shovel to about 15 cm depth

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

and 8 cm diameter. Soil from each sample was thoroughly mixed before taking a subsample for

analysis.

Appendix 1.2

Analysis of microplastic in wastewater and soil

Sample preparation and concentration for micro-FT-IR image analysis

The sample preparation and concentration has the objective to create a concentrate of plastic

particles suitable for the subsequent FT-IR analysis. While the sample preparation methods

differ for the matrixes investigated, the endpoint is always the same, namely a concentration of

particles suspended in 5 mL of ethanol.

Preparation of raw wastewater

First, 1 L of raw wastewater was pre-sieved using a 500 µm mesh size as this study only focus-

es on particles smaller than 500 µm. However, due to the strong adherence of plastic particles

to other organic or inorganic particles, 1 mL of 150 g/L sodium dodecyl sulphate (SDS) was

added to the wastewater sample before the wet pre-sieving. This was done to ensure that mi-

croplastic particles in all size ranges were detached from other particles and were sieved out

into their proper size range.

Figure 19. Material from oxidized wastewater filtered onto a 10 µm steel mesh. The top

row shows the result without prior use of cellulase incubation. The bottom row shows

the effect of pre-treating with cellulase.

Wastewater contains a large fraction of cellulose fibres, which cause problems in the micro-FT-

IR analysis. To eliminate these fibres, a cellulose digesting enzyme (cellulase from an Aspergil-

lus species) was added to 200 mL of the pre-sieved wastewater (Figure 19). The samples were

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

incubated for 48 hours at 40°C. Afterwards, the hydrolysed samples were oxidised by adding

50% hydrogen peroxide (H

) to an initial concentration of 330 g/L and adding a catalyst to

enhance the reaction. The oxidised samples were wet sieved into two fractions: larger than and

smaller than 80 µm.

Particles were removed from the sieves by ultrasonic treatment and scraping and gathered into

filtered demineralised water (filter size 0.8 µm) containing 0.15 g/L SDS. The liquids of each

fraction were filtered separately through a 10 µm stainless steel filter mesh (a mesh as shown in

Figure 19). After filtration the filters were immediately put into ethanol, sonicated and scraped.

The liquid was transferred into a glass vial and the beaker was flushed 3 times with ethanol.

After the final flushing, ethanol was completely evaporated from the vial. Then 5 mL ethanol

was added and the vial was sonicated. Particles stick very much to the glass when they are dry.

Therefore, they need to be removed from the glass walls and into the ethanol. The ethanol

containing the particles was evaporated to 5 mL in volume, stored, and at a later time trans-

ferred to a Zinc Selenide transmission window or MirrIR slide as described in Appendix 1.3.

Preparation of treated wastewater samples

As described previously, the sampling of treated wastewater resulted in 3 Ø47 mm steel filter

meshes of 10 µm pore size. These filter where hydrolysed and oxidized similar to the raw

wastewater samples, resulting in two fractions of particles: larger than and smaller than 80 µm

suspended in 5 mL of ethanol.

Preparation of sludge samples

Sludge samples were prepared by taking 0.1 g dry matter and suspending it in 50 mL of milliQ

water. The sample was then treated similar to the raw wastewater, but for correspondingly

modified sample volumes.

Preparation of soil samples

Soil contains a large inorganic fraction (grit, sand, silt and clay) which has to be separated from

the organic fraction. However, due to the strong tendency of microplastics to adhere to other

particles, samples were first treated with a surfactant (SDS) solution. A soil sample of 50 g was

suspended into approx. 1 L filtered demineralised water with 0.15 g/L of SDS. In order to force

the microplastic to separate from the soil, it was slightly agitated before the suspension was

wet-sieved with several litres of SDS solution into fractions 10-80 µm and 80-500 µm. From

here on, each fraction was treated identical, but kept separated for later FT-IR analysis. The

particles were retrieved from the sieves in the same manner as described for the wastewater

treatments, except that the filters were sonicated into zinc chloride solution and not into ethanol.

Each soil fraction was transferred to a straight glass funnel with stopcock, for separation of the

-1

inorganic fraction (Figure 20). This was done by gravimetrical separation using a 1.7 kg L zinc

chloride solution made by dissolution in 0.8 µm filtered demineralised water. The soil sample

suspended in the zinc chloride solution was agitated for approximately 15 minutes by aeration

from below. After the aeration was turned off, organic material sticking to the inside of the glass

was flushed into the bulk liquid with fresh zinc chloride solution. The column was left for sedi-

mentation/floatation and the approx. top 8 cm of liquid column was drained (approx. 100 mL)

through a side port. The zinc chloride solution was resupplied and the floatation sequence was

repeated twice. For the first two flotations, the column was left for sedimentation/floatation for 2

hours, while it for the third floatation was left over night.

The mixed liquid obtained from the floatation sequence was then filtered over a 10 µm steel

mesh filter. Here upon the material from the filters were oxidised as described for treated

wastewater. For soil samples, the material was not treated with cellulose prior to oxidation. After

removing the organic matter, the sample was again filtered over a 10 µm steel mesh and the

particles transferred from the filter into ethanol using a combination of washing, scraping and

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

ultrasonic treatment. The ethanol containing the particles was evaporated to 5 mL, stored, and

at a later time transferred to the appropriate FT-IR windows.

Figure 20. Gravimetrical separation funnel. The blue line supplied the compressed air

(CA) for aeration. Between the tubing connecting the glass funnel to the CA, an air filter

was placed to reduce contamination. At the 600 mL mark there was a side port with a

cork stop, to allow for draining the fluid containing low density particles

Micro-FT-IR spectroscopy for sample imaging

The equipment used to identify and quantify microplastic was a Fourier Transform Infrared (FT-

IR) system, consisting of an Agilent Cary 620 FT-IR microscope combined with an Agilent Cary

670 FT-IR spectrometer (Figure 21). The microscope has a 128 x 128 pixel Focal Plane Array

(FPA). The equipment can operate in reflection, transmission and ATR (Attenuated Total Re-

flectance) mode. The equipment has 4, 15 and 25x objectives, allowing pixel sizes on the FPA

(in transmission and reflection mode) of 20.6, 5.5, and 3.3 µm, respectively. It can furthermore

operate in a ‘high magnification mode’ that allows for a 5 times finer pixel resolution, i.e. the

lowest pixel resolution with which the equipment can operate is 0.66 µm. In ATR mode the

equipment can operate at 1.1 µm pixel resolution.

Appendix 1.3

Spectral analysis

The spectral analysis is done by interpreting at which wavenumbers various chemical bonds

absorb energy. An example of such interpretation is given below for 3 different particles: A

natural cellulose particle, a polyamide/nylon particle and a polyethylene particle (Table 8).

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Table 8. Examples of analysis of spectra from particles

Discernible Peaks

3360

2915 & 2849 split peak

1746

1635 with shoulder at

1604 & 1539

1428 with shoulder at

1459

1370 + 1336 + 1317 peak

with shoulder at 1397

1281 + 1250

1205

1052 & 1031 split peak

shoulders at 1160 + 897 +

863

3295

2921 & 2851 with shoul-

der at 2954 & 2868

1742

1648 & 1547 with split at

1537

1464 & 1445 split peak

1400 with 3 additional

peaks

1245

1108

1015

2913 & 2847 split peak

1462

Peak Assignment

O-H stretch

Aliphatic C-H stretch. Asym & symm

methylene stretch

C=O stretch

C=O stretch of Amide I & C-N stretch

& N-H bend of Amide II

Asymm deformation methyl

Methyl symm deformation + other

peaks

Poss. C-O stretch

Particle Definition

Cellulose

Confidence

Probably

O-H stretch with clear N-H stretch

Aliphatic C-H stretch. Asym & symm

stretch methylene + asym stretch

methyl

C=O stretch

C=O stretch of Amide I & C-N stretch

& N-H bend of Amide II

Asymm & symm deformation of

methyl + methylene

C-H deformation

C-O stretch

Polyamide or Nylon

High

Aliphatic C-H stretch. Asym & symm

C-H stretch of methylene

Asym C-H deformation of methyl

groups

Polyethylene

High

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 21. Micro-FT-IR spectrometer and imaging system

The micro-FT-IR spectrometer and imaging system creates a visual image and infrared (IR)

image. The images are made up of tiles depending on the chosen resolution of the imaging

system, and the tiles can be combined to a mosaic. The IR tiles consist of a full IR spectrum per

pixel and is visualised by focusing on one IR wavelength at a time. An example of such a com-

bined view is shows in Figure 22. On the top left hand side the visual image is shown; the top

right shows the IR data image of the same area, showing pixels that transmitted less IR light at

-1

the chosen wavenumber peak at 2326 cm . The lower hand image is the full IR spectrum of the

pixel indicated by the cross on the upper images. For identification of the particle’s material, its

full IR spectrum is interpreted by spectral analysis.

Figure 22. Scrapes of plastic particles including material from blue-cab bottles. The par-

ticles are placed on a gold reflection window

In some ways, the simplest FT-IR operation mode is reflection. The sample is transferred to a

Kevley MirrIR microscope slide, made especially for IR reflection scanning. The IR light is re-

flected off the particles on the slide and as the particles only reflect part of the wavelengths

transmitted, their material can be identified from the resulting reflection spectrum. However, due

to scattering during the reflection, this mode has a low signal strength compared to the signal

strength in other modes (transmission, ATR). Furthermore, small particles amplify this effect,

meaning that reflection mode is only an adequate option for comparatively large microplastic

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

particles. For the equipment used in the present study and for the type of sample analysed,

reflection mode is used for particles from 80 µm and up. Two size fractions of samples are

hence scanned: 10-80 µm and 80-500 µm.

For smaller particle sizes, transmission mode or ATR mode is required in order to acquire better

IR spectra. Of these methods, transmission mode allows more freedom in scanning. For trans-

mission mode, the particles are transferred to an IR transparent microscope window (Zinc

Selenide) and the transmission of IR light through the particles is measured. This method de-

mands that the particles are sufficiently thin to allow IR transmission with good signal strength.

This means that transmission mode is good for smaller particle sizes, but not always adequate

for larger plastic particles. In both transmission and reflection mode, the system is able to au-

tomatically create mosaics of larger sample areas, by stitching together adjacent tiles.

ATR is generally considered the most sensitive of the methods and works by bringing an ATR

crystal into physical contact with the sample. While this method does not demand transferring

the sample to a transmission or reflection window, it has the drawback that when pulling up the

ATR crystal from the sample, some particles tend to stick to the crystal. As a result, scans can

only be taken once and the crystal has to be cleaned between each scan. Because of this ATR

mode cannot be used to create an automated mosaic of a larger sample area.

One important issue when operating a micro-FT-IR spectroscopy system is interference with IR-

absorbing constituents of the atmosphere. While this is not an issue when applying ATR (as the

ATR crystal is brought in contact with the sample and the IR light hence does not have to pene-

trate through atmospheric air), it is a significant issue when operating in transmission or reflec-

tion mode. Here the modus operandi is that first one takes a background spectrum of the inter-

fering air. Then one takes a spectrum of the sample, from which the background then is sub-

tracted. The main constituents interfering with the measurement are CO

and H

O gases. Es-

pecially the latter absorbs IR light in a region of the spectrum that is critical for the identification

of plastic polymers. It is hence critical to ensure a dry and especially highly stable atmosphere

between the objective and the sample window. Not doing so creates spectra that cannot be

reliably interpreted. It hence drastically increases the risk of obtaining false positives or false

negatives in the spectral analysis as peaks from water vapour interference can be interpreted

as peaks from plastic materials.

For analysis in reflection mode, a MirrIR microscope window has been prepared by fixing a

steel washer with an inner diameter of 10 mm on the window. A well-defined amount of sample

solution (particles suspended in ethanol) is then transferred to the restricted area. After the

ethanol has evaporated, the window is checked in a traditional light microscope to ensure that

the window is adequately populated by particles. If the slide is insufficiently populated, addition-

al particle suspension is added till the window is deemed sufficiently populated. Here after the

slide is placed the microscope stage.

For transmission mode, pre-treated samples are transferred to a transmission window. The

windows used are Ø13 mm Zinc Selenide (ZnSe) windows mounted in a compression cell

(Figure 23). The population of the window is ensured similar to the reflection windows. The

compression cell has a clear aperture of 10 mm, in which the sample is deposited. The sample

window is taken out of the compression cell, placed on a metal disk which is then placed on the

microscope stage and analysed.

The area of the window scanned (i.e. the number of tiles scanned) depends on how rich the

sample is on plastic particles. Based on experience for the various types of samples, the num-

ber of tiles scanned is set for each sample type. The number of actually analysed tiles differs

with population density of plastics on the windows.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

The micro-FT-IR system applied in the study is equipped with a 128x128 pixel FPA (focal plane

array); however, due to technical problems with the equipment, the majority of the scanning had

to be done using only a quarter of this size, namely 64x64 pixel FPA. I.e. to achieve the same

area operating in 64x64 FPA mode, 4 times as many tiles are needed to cover the same area

(and unfortunately also 4 times the scanning time). Hence when analysing raw wastewater,

treated wastewater, sludge, and soil samples the predefined number of tiles scanned were:



Transmission mode, 80-500 µm

Raw wastewater: 64x64 pixel FPA: 16 tiles; 128x128 FPA: 4 tiles

Treated wastewater: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles

Sludge: 64x64 pixel FPA: 16 tiles; 128x128 FPA: 4 tiles

Soil: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles



Reflection mode, 10-80 µm

Raw wastewater: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles

Treated wastewater: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles

Sludge: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles

Soil: 64x64 pixel FPA: 32 tiles; 128x128 FPA: 8 tiles

The scan time for one tile depends on the actual sample, but is between 5 and 8 minutes. For

32 tiles the resulting total scan times including startup and finalization of the scans is hence 3-5

hours per sample.

Figure 23. Compression cell for transmission mode imaging. The cell is only used for

evaporating the sample onto a well-defined area of the transmission window, and, de-

spite of its name, never actually used to compress two windows.

Spectral analysis of particles

As discussed previously, one should not simply compare measured particle spectra to refer-

ence spectra as this will lead to an increased risk of both false positives and false negatives. A

better approach is to perform an initial screening of spectra for materials which could be plastic

– i.e. rapidly rejecting for example mineral particles, which in most samples constitute the ma-

jority of the particles on a window. The candidate particles are then, one by one, analysed for

the material composition by identifying all assignable peaks in their spectra. Figure 24 illustrates

this principle.

First all particles that absorb IR light at the wavelengths marked by blue in the lower image are

highlighted in the image analysis software. Then particles showing significant absorption are

selected, for example the one indicated in the two upper images. The analyst goes through all

the relevant data starting with the first collected tile and moving right and down. The peaks in

the spectra of the highlighted particles are analysed, providing an indication of the material type

(in this case PE). In some cases a spectrum contains peaks characteristic to for example poly-

ethylene, while also containing other peaks. These are all noted down, and when a determina-

tion is made, it receives a confidence scale which rates from Maybe to Possible, Likely and

High. Maybe is a relatively low quality spectra but peaks are still attributable, Possible relates to

not all peaks being found or a few too many (for example from interference from scattering).

Likely means all peaks are present and the spectral quality is OK, some interference peaks may

be present or uncertainty given due to the large permutations of possible copolymers. High is

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

only attributed when a spectrum is of such good quality and clarity that the analyst could not

possibly assign it as anything else. When the analyst is in doubt, a second opinion is asked.

Figure 24. Searching for plastic particles on a 700x2800 μm section of an IR transmission

window

A zoom on the particle shows its shape and size as 40x45 μm (Figure 25). The image is of

course only 2-dimensional, and the thickness of the particle must hence be estimated. In the

present study, it is as a general rule assumed that the thickness of a particle is 60% of its short-

est axis. This assumption is not substantiated by evidence, and hence a source of uncertainty.

It does, though, seem reasonable to assume that particles orient themselves on the window

with the smallest dimension pointing upwards.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 25. A zoom on the PE particle identified in Figure 24

Appendix 1.4

Calculation of particle volume and mass

For each particle the longest and shortest diameter was noted down from the light microscopy

image. For calculation of the particle volume it was assumed that particles were ellipsoidal and

the volume calculated by applying the two determined diameters as two of the ellipsoid diame-

ters. The third diameter of the ellipsoid was set to 60% of the smallest determined diameter.

The plastic mass was found as this volume multiplied by reference densities of the plastic mate-

rial (Table 9). With the above discussed assumptions and limitations, the identified particle of

Figure 25 is estimated to have a mass of 22 ng.

Table 9. Plastic materials and their densities. Sources: British Plastics Federation

(bpf.co.uk) and Prospector (ulprospector.com).

Material

PAM/Nylons

HDPE

LDPE

PTFE

Polyesters

Polystyrene

Polyvinyl Chloride

Polyurethane

(very variable in density)

TPU foam

Generic PUR

PUR-Ester

Generic PUR (MDI/TDI)

PUR-Glasfiber

SBR

Zinc Stearate coated particles

Abbreviation

Density range

[g/cm

]

1.13-1.41

0.905

0.944-0.965

0.917-0.930

2.14-2.19

1.05

1.03-1.06

1.38

0.4-1.0

0.94-1.11

1.15-1.27

0.0451 - 1.25

1.26-1.67

1.03 - 1.05

1.08

1.069

Density average

[g/cm

]

1.27

0.905

0.94

0.929

2.155

1.37

1.05

1.045

1.38

1.025

1.21

GPPS

HIPS

ABS

ASA

1.04

* Assuming that Zinc stearate is coated on most popular household appliance plastics such as

PP, PE & PS

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Appendix 2.

Method

validation,

uncertainties,

and detection

limits

Appendix 2.1

Recovery of plastic particles

In the present study we have approached the validity of the sample preparation method by

recovery of microplastic particles. We spiked known plastic particles to raw wastewater and

quantified the recovery in terms of numbers of particles. The validation was done by spiking the

raw wastewater with 3 types of plastic particles before any pretreatment: Red 100 µm spherical

polystyrene particles (Sigma Aldrich, product no. 56969), light-green high-density polyethylene

(HDPE) particles, size 80-150 µm, made from a HDPE water bottle, and red styrene butadiene

rubber (SBR) particles made from a SBR sheet (Figure 26).

Instead of analysing only part of a sample for microplastic applying FT-IR, the complete sample

was analysed by light-microscopy. This was made possible by the highly characteristic colours

and shapes of the materials used for spiking. The number of plastic particles added to

wastewater samples was quantified by counting under a light microscope. A known number of

particles were then added to 200 mL of raw wastewater, upon which the sample underwent the

same preparation steps as presented in Appendix 1.2. The complete sample was then counted

by means of light-microscopy.

Figure 26. Microscope images of the plastic particles used in this study for method vali-

dation.

The obtained recoveries when sonicating the 10 μm filters once (i.e. the transfer of particles

from 10 μm filter meshes to ethanol) are shown in Figure 27. On average the recovery for poly-

styrene was approx. 85% while HDPE particles were somewhat lower. The recovery of SBR

particles were above 100% which most likely was due to breakup of particles during sample

preparation.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 27. Average recovery for each type of plastic from wastewater samples when

sonicating filters once.

The recovery presented in Figure 27 led to a scrutinizing of the method and it turned out that

even better recovery could be obtained by repeating the sonication (i.e. the transfer of particles

from the filters to the ethanol). Hereby a close to 100% recovery could be obtained. The proce-

dures for sample preparation were hence adjusted accordingly.

Appendix 2.2

Uncertainties

The above presented tests on recovery show some of the issues of quantifying microplastic in

environmental samples: Not all plastic particles behave the same, and plastic particles might

actually break down during the sample preparation. Nevertheless, the uncertainty of the sample

preparation method seems to be rather good – at least for the plastic particles used to spike the

natural wastewater samples.

Sample preparation is, though, not the only uncertainty to be faced. A second uncertainty lies in

the fact that not all of a sample can be analysed by micro-FT-IR imaging. To do so would re-

quire unrealistic scanning and interpretation times (weeks of scanning). The uncertainty of se-

lecting a sub-area of slide compared to the whole slide has not been assessed in the present

study, but is believed not to introduce a systematic error and hence equal out when analysing

many samples for determining average microplastic contents.

A third uncertainty lies in the spectral analysis of the FT-IR images. The spectra of particles

from natural samples seldom look like the reference spectra of the pure materials. There hence

lies uncertainty in analysing the spectra, and the analysis furthermore becomes to some degree

subjective. A similar issue is well-known from biological analysis, for example when identifying

and counting algae in a water body. Again similar hereto, this uncertainty can be minimized by

increasing the time used to analyse and understand each and every spectrum. Quantifying this

uncertainty is, though, not straight forward.

Appendix 2.3

Detection limits

Defining detection limits for microplastic particles is not straight forward, and one has to consid-

er whether detection limits are to be defined in terms of number of particles or in terms of parti-

cle mass. In the present study we have somewhat circumvented the issue of detection limits as

we simply have identified plastic particles until at least 10 plastic particles within a certain size

range were found. We have, though, limited ourselves to a maximum number of tiles analyzed

to keep the scanning and interpretation times within limits. Detection limits were hence variable.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

The detection limit of the method depends on the fraction of the sample scanned by the FT-IR

imaging system. We have scanned till at least 10 plastic particles had been positively identified,

or till we had scanned an area of at least 4 mm on the microscope slide. The microscope slide

was circular with a radius of 5 mm and hence had an area of 78 mm .



For wastewater samples, particles from 0.2 L were concentrated into 5 mL of ethanol.

Of this, at least 0.2 mL was deposited on the scanned slide. Assuming that micro-

plastic particles are of size 20 μm, this corresponds to a detection limit of better than 4

μg/L.



For treated wastewater, particles from between 4.1 and 81.5 L were concentrated into

5 mL of ethanol. Of this, at least 0.2 mL was deposited on the scanned slide. Assum-

ing that microplastic particles are of size 20 μm, this corresponds to a detection limit of

better than 0.20 μg/L.



For sludge, particles from 0.1 g were concentrated into 5 mL of ethanol. Of this, at

least 0.1 mL was deposited on the scanned slide. Assuming that microplastic particles

are of size 20 μm, this corresponds to a detection limit of better than 20 μg/g.



For soil, particles from 50 g were concentrated 5 mL of ethanol. Of this, at least 0.1 mL

was deposited on the scanned slide. Assuming that microplastic particles are of size

20 μm, this corresponds to a detection limit of better than 0.04 μg/g

Appendix 2.4

Identification of tire rubber

During the course of the project there have been voiced concerns whether rubber from car tires

can be identified by the applied FT-IR technique and spectral analysis. The concern was that

the carbon black content of car tires would absorb basically all IR light and render spectral

analysis impossible. To test this issue, a series of tests were conducted on car tire rubber.

Car tires are an amalgam of several compounds, whereof the main components are copolymer

rubbers and fillers. The rubbers used in car tires are often a combination of natural rubber (poly-

isoprene), styrene rubber and butadiene rubber. The exact composition changes with the tire

producer and the wished use for the tire. Typically, the amount of synthetic rubber will decrease

with an increasing need for heat resistance, therefore tires for busses, trucks and aircrafts will

contain higher natural rubber contents than those of passenger cars. In general, the rubber in a

tire will consist of 40-60% of its total mass, while the next large contribution will come from

reinforcing fillers (20-50%). The most common reinforcing fillers are carbon black and amor-

phous silica. Due to low cost, carbon black was the most common filler in the automotive indus-

try for a long time, but slowly amorphous silica becomes more widely used as technological

advances are made.

The different components of rubber copolymers have specific infrared absorbance spectra, as

they all absorb some of the infrared light. This means that a spectrum taken from car tire frag-

ments will also show a combination of these components. This spectrum will naturally not look

like only that of the rubber but will provide a type of fingerprint for each car tire dependent on its

composition. In Figure 28 four spectra are displayed which were acquired from four different car

tire brands, and of which two are summer tires and two winter tires. As can be seen, even

though the exact composition may be different, the spectra are regardless very similar to each

other, since most tires are made from the same type of styrene and butadiene copolymer (SBR)

and similar reinforcing fillers.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 28. FT-IR-ATR spectra taken from 4 different brands of car tires, of which 2 were

winter and 2 were summer tires. The spectra are slightly offset from each other to be

able to distinguish between them.

Car tires do contain carbon black and pure carbon black does indeed absorb all light at all

wavelengths, and a pure carbon black sample can hence not be analysed by the applied tech-

nique. However, samples having carbon black as one constituent among others can be meas-

ured using FT-IR as shown also by other researchers (Murakami, 2009). The carbon black will

have an impact on the provided spectra, but the change has been shown to be a predictable

one, and can be accounted for in spectral analysis and interpretation. Dissimilarly to what was

seen in the Shimadzu study, the tire samples analysed in the present study did not show prob-

lems related to carbon black.

In ensuring that identification of car tire fragments is possible, even after treatment, an analysis

has been made of fragments treated with possible affecting agents used in the sample prepara-

tion method. The two processes considered possibly detrimental to the FT-IR spectral outcome

are the oxidation and ultra-sonication processes. Therefore, a small triplicate study was done of

one of the car tires for which spectra were acquired, after having undergone treatment. Figure

29 shows these respective spectra before treatment, after oxidation only and after oxidation and

ultra-sonication. It seems that especially the ultra-sonication may result in some slight alteration

of the car tire spectra. However, it is mainly in absorbance intensity, which is already variable

due to differences in composition between different tire brands. The spectral analysis as such

was hence not affected by the treatment methods. Hence there is no reason to believe that the

treatment method of the present study affected identification of car tire particles in itself.

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Figure 29. Spectra collected from a triplicate treatment test to assess the impact of sam-

ple treatment on the infrared spectra of a given car tire rubber composite.

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Appendix 2.5

Data overview

Overview of raw data. Samples for wastewater: WW, treated wastewater: TW, Sludge: Sludge. The following number indicates the plant number. Soils

are indicated by Soil with sludge and Soil without sludge. The analysed volume refers to the corresponding volume analysed at the FT-IR imaging sys-

tem. Number of particles describes the plastic particles identified. Min and max dimensions are the smallest and largest dimensions of any particle.

Plant number

Analyzed

volume

[mL]

5.93E-01

3.66E-01

8.58E-01

2.46E-01

9.16E-02

1.01E+00

2.75E-01

7.91E-01

5.30E-01

3.53E-01

1.37E-01

3.03E-01

2.90E-01

2.40E-01

Number of

particles

Mass of

particles

[mg]

7.18E-05

8.00E-04

3.94E-04

9.92E-04

3.36E-04

3.72E-03

2.36E-03

4.19E-04

2.08E-03

8.88E-04

2.62E-04

4.37E-04

3.80E-04

1.83E-03

Min dimen-

sion

[μm]

Max dimen-

sion

[μm]

100

120

107

215

100

190

125

100

No concentration

[no/L]

12711

98443

19148

237704

760903

85837

52273

69624

70092

85177

141962

16507

144063

442393

Mass concentra-

tion

[mg/g]

0.22

4.86

1.14

13.13

6.05

11.13

15.18

1.35

6.80

6.29

2.99

1.44

5.47

29.55

Plastic types

WW1

WW 2

WW 3

WW 4

WW 4, only parti-

cles below 80 μm

WW 5

WW 5, only parti-

cles below 80 μm

WW 6

WW 7

WW 8

WW 8, only parti-

cles below 80 μm

WW 9, only parti-

cles below 80 μm

WW 9

WW 10

PE-co-polymer, PA/NYlon

PA/NYlon

PE-co-polymer, PA/NYlon

PE-co-polymer, PA/NYlon, LDPE

PE, PA/NYlon, LDPE, PVC, PP

PA/NYlon

PE-co-polymer, PA/NYlon

PE-co-polymer, PA/NYlon, Zn-

stearate coated particle

PE, PA/NYlon

PE, PE-co-polymer, PA/NYlon

PA/NYlon

PA/NYlon, PE co-polymer, PP co-

polymer, Zn-stearate coated particle

Plastic types

Plant number

Analyzed

volume

[mL]

3.52E+01

1.44E+01

1.26E+02

6.76E+01

2.49E+01

Number of

particles

Mass of

particles

[mg]

1.54E-03

3.98E-04

3.30E-04

5.76E-04

3.79E-04

Min dimen-

sion

[μm]

Max dimen-

sion

[μm]

120

100

270

No concentration

[no/L]

1801

4254

6933

503

6053

Mass concentra-

tion

[mg/g]

0.1336

0.0930

0.0056

0.0170

0.0325

TW1

TW2

TW3

TW4

TW5

PE, PA/NYlon, PE co-polymer

PP, PA/NYlon, Zn-stearate coated

particle

PA/NYlon, PE-co-polymer

PA/NYlon, PE-co-polymer, PE

Environmental Protection Agency / Microplastics in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

TW6

TW7

TW8

TW9

TW10

2.28E+01

1.71E+02

2.78E+02

1.65E+02

1.37E+02

2.71E-04

3.74E-04

3.66E-04

2.39E-04

9.94E-04

120

140

406

10389

6650

8716

4695

8419

0.0384

0.0043

0.0028

0.0021

0.0145

PA/NYlon, PP

PA/NYlon, PE co-polymer

PA/NYlon, PE

PA/NYlon, PE co-polymer

PA/NYlon, PP

Plant number

Analyzed

volume

[mg]

1.70E-01

1.34E-01

1.26E-01

2.84E-01

1.34E-01

Number of

particles

Mass of

particles

[mg]

4.13E-04

2.74E-04

3.27E-04

9.00E-05

8.17E-05

Min dimen-

sion

[μm]

Max dimen-

sion

[μm]

No concentration

[no/g]

79235

413780

158469

17608

174976

Mass concentra-

tion

[mg/g]

7.23

6.45

6.48

0.40

2.02

Plastic types

Sludge 1

Sludge 2

Sludge 3

Sludge 4

Sludge 5

PE-co-polymer, PP

PE-co-polymer, PA/NYlon

PE, PA/NYlon

PE-co-polymer, PA/NYlon

Plant number

Analyzed

volume

[mg]

5.28E+01

5.60E+01

3.03E+02

1.14E+02

1.70E+02

2.27E+01

8.20E+01

1.53E+02

3.53E+01

8.52E+01

Number of

particles

Mass of

particles

[mg]

0.00E+00

1.62E-04

7.26E-04

4.43E-04

2.41E-04

1.69E-04

5.73E-04

4.14E-04

2.29E-03

4.83E-06

Min dimen-

sion

[μm]

Max dimen-

sion

[μm]

110

130

120

150

120

245

No concentration

[no/g]

143

165

528

145

436

Mass concentra-

tion

[mg/g]

0.0000

0.0058

0.0165

0.0076

0.0014

0.0149

0.0117

0.0044

0.2243

0.0001

Plastic types

Soil with sludge 1

Soil with sludge 2

Soil with sludge 3

Soil with sludge 4

Soil with sludge 5

Soil without

sludge 1

Soil without

sludge 2

Soil without

sludge 3

Soil without

sludge 4

Soil without

sludge 5

5.28E+01

PP, PE co-polymer, PA/Nylon

PE co-polymer, PA/Nylon

PE co-polymer, PA/Nylon, PE

PE co-polymer

PE co-polymer, PE, PA/Nylon

PE co-polymer, PE

PE co-polymer, PP

PE co-polymer, PA/NYlon

PA/NYlon

Environmental Protection Agency / Microplastic in Danish wastewater

MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate MOF, Alm.del - 2016-17 - Bilag 303: Rapporten Microplastic in Danish wastewater - Sources, occurrences and fate

Microplastic in Danish wastewater – Sources, occurrences and fate

The objectives of the present study were to evaluate the role of Danish wastewater

treatment plants (WWTPs) in the emission of microplastic to the environment in terms

of amounts and types of plastic polymers emitted and if possible, to evaluate which

sources these plastic polymers could originate from.

Samples from 10 WWTPs (wastewater, inlet and outlet), sludge from 5 of these

plants, and 10 farmlands soils (5 soils that had received sludge as fertilizer and 5 that

had not) were analysed for the occurrences of microplastic with the currently most

advanced method available for microplastic investigations (Fourier Transformed

Infrared Spectroscopy imaging applying a Focal Plane Array). This method allows

both determination of the microplastic concentrations in the samples and identifica-

tion of the type of plastic polymer of each microplastic particle.

Environmental

Protection Agency

Strandgade 29

DK-1401 København K