Emission inventories for priority substances at catchment levels: Solving the PAH source conundrum with an array of in‐stream tools

T. Gallé Michael Bayerle, Denis Pittois, Andreas Krein Institue of Science and Technology Contact: [email protected] BAD CHEMICAL STATUS PAH EQS exceedance most common reason

• All surveillance sites in Luxembourg exceed EQS for high-molecular weight PAH

• Same trends in the neighboring regions (Rhine-Mosel Commission)

• PAH often considered as ubiquitous with important atmospheric immission

• Fatalistic attitude upon improving the situation (little concrete measures in RBMP)

• Scarce efforts to investigate spatial differentitation and sources more thoroughly

2 PAH SOURCES AND DYNAMICS What the literature says

• Exports from catchments > current atmospheric deposition (urban areas)

• Street deposits major source

• Soils often secondary source

• Accumulation in soils in vicinity of traffic

• Building up of stocks in sewers (first flushs)

• Role of combustion derived carbonaceous particles

• Contaminated industrial sites (historical)

3 EMISSION INVENTORIES One scheme for all compounds?

4 • Is this adapted to secondary pollutants with diffuse sources and a strong affinity for solids? SUBSTANCE FLOW ANALYSIS Regionalized emission balances

Susp. matter [mg/L] 500 • Substance flow Ettelbruck analysis in different 400

catchment 300

• Establishment of Q-C Data: Data1_B 200 Model: Allometric1 Equation: y = a*x^b relationships Weighting: y No weighting

Chi^2/DoF = 1240.40378

Susp. matter [mg/L] matter Susp. 100 R^2 = 0.76735

• Calculation of a 0.92937 ±0.34971 b 1.32497 ±0.09034 catchment loads 0 • Characterization of 0 20 40 60 80 100 120 Discharge [m3/s]

pollution level in 7000 Ettelbruck different hydrological 6000

situations 5000 y = a*x^b R^2 = 0.98238

4000 a 0.02843 ±0.03588 • Contribution of b 3.11534 ±0.32406 3000 Sum PAH

WWTPs and urban [ng/l] PAH  runoff 2000 1000

0

-1000 0 5 10 15 20 25 30 35 40 45 50 55

3 5 Discharge [m /s] SUSPENDED MATTER POLLUTION Source discrimination

35 PAH solid > 15 mg/L Susp M Sum PAH solid > 15 mg/L susp. m Sum PAH solid > 15 mg/L susp. matter 30 Steinsel 30 Ettelbruck 30 Hunnebour 25 25

20 20

20

15 15

10

PAH solidPAH [mg/kg] PAH solid [mg/kg]

PAH solidPAH [mg/kg] 

 10 10  5 5 0 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 0 0 100 200 300 400 Susp. M. [mg/l] Susp. M. [mg/l] Susp. M. [mg/l]

• Solid contamination levels allow for objective comparison in different hydrological situations and catchments (SPM as main carrier) • High levels of SPM indicate strong catchment wide erosion and background levels of (alluvial) contamination 6 CATCHMENT BALANCES Yearly variability - uncertainties

Daily discharge [mM3] Daily discharge [Mm3] 25 Daily average PAH solid [mg/kg] Ettelbruck 250 25 Ettelbruck 250 Daily average PAH solid [mg/kg] DAily PAH 16 load [kg] Daily PAH 16 load [kg] 20 200 20 200

]

3

] 3 15 150 15 150 2002-2003 2003-2004

3 Discharge 401 Mm3 Discharge 210 Mm 10 PAH 16 load 123 kg 100 10 PAH 16 load 824 kg 100

Average PAH 16 conc. 2.05 g/L Average PAH 16 conc. 0.58 g/L Daily 16loadPAH [kg]

Daily loadPAH 16 [kg]

Dailydischarge [Mm

Dailydischarge [Mm 5 50 5 50

Average solid [mg/kg] PAH 16 Average solid 16[mg/kg] PAH

0 0 0 0 01/04/2003 01/08/2003 01/12/2003 01/04/2004 01/04/2002 01/08/2002 01/12/2002 01/04/2003

Steinsel Ettelbrück Steinsel Ettelbrück 2002-2003 2002-2003 2003-2004 2003-2004 Yearly discharge [Mm3] 183 401 99 210 Yearly load PAH [kg] 241 824 56 123 Yearly average 1.32 2.05 0.58 0.58 concentration PAH [µg/L] Yearly average 95 160 33 37 concentration SPM [mg/L]

• Yearly loads are governed by SPM yield – read: hydrological events 7 • SPM pollution levels is the more objective measure CONTRIBUTION BY URBAN AREAS WWTPs vs runoff pollution

500 • Contribution of WWTPs: WWTP outlets Schifflange medians of measurement 400 Bettembourg Bonnevoie Beggen campaigns 300 Mersch

• Contribution of surface runoff 200 PAH 15PAH[ng/l] • Combined sewers 100 Difference between 0 A B C D E effective precipitation 25 Schifflange Bettembourg Bonnevoie Beggen Mersch on impervious surfaces WWTP outlet SPM and discharge of WWTPs 20 • Separative sewers: Effective 15 precipitation 10

• Median concentrations: solidPAH [mg/kg] 5

measurements/literature 0 A B C D E Schifflange Bettembourg Bonnevoie Beggen Mersch Beringen 8 CONTRIBUTION PATTERNS Runoff as the main source?

Alzette by Difference Alzette by Difference Steinsel PAH loads [kg/y] WWTP Steinsel discharge [Mm3/y] WWTP 2002-2003 CSO 2002-2003 CSO Stormwater Stormwater

5.52 (3.02%) 149.93 (81.93%) 31.9 (17.45%) 15.9 (8.69%)

142 (77.68%)

5.7 (3.12%) 11.65 (6.37%)

3.2 (1.75%)

Alzette by Difference Steinsel Cu loads [kg/y] tot Alzette by Difference WWTP WWTP Steinsel PAH loads [kg/y] 2002-2003 CSO CSO 2003-2004 Stormwater Stormwater 138 (12.59%) 249 (22.72%) 5.34 (9.54%) 12.56 (22.43%)

8.76 (15.64%) 133 (12.14%)

576 (52.55%) 29.34 (52.39%) • The contribution of WWTPs and urban runoff is much smaller for PAH than for metals

9 • The contribution is largely dependent on the hydrological season (wet year vs. dry year) FLOOD EVENTS Source mobilisation and transport dynamics

• Follow-up project on urban runoff through in- stream balances • 3 triggerd autosamplers in a longitudinal profile from industrial region to strongly urbanized segments • Event based balancing and peak analysis in flood waves

• Can we depict fast urban runoff contributions in chemographs?

10 EVENT MEAN CONCENTRATIONS Fingerprinitng the sources Hesperange Pfaffenthal Livange 200 150 250

Cu solid [mg/kg] Model: Allometric1 Model: Allometric1 200 150 Equation: Equation: y = 73,93306*x^-0,34051 y = 63,96336*x^-0,23224 100 R^2 = 0.70805 R^2 = 0.87077 150 100

100 50

50 EMC Cu solid [mg/kg]

EMC Cu solid [mg/kg] 50

0 Event mean concentration [mg/kg] 0 0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 3 3 Q tot [Mm ] 3 Q tot [Mm ] Q tot [Mm ]

Livange Hesperange Pfaffenthal 40 40 40

Data: EMC_J Sum 16 PAH [mg/kg] Model: Allometric1 Sum 16 PAH [mg/kg] Data: EMC_J Equation: Model: Allometric1 y = a*x^b Sum 16 PAH [mg/kg] Equation: Weighting: y = a*x^b y No weighting Weighting:

Chi^2/DoF = 13.59298 30 30 y No weighting 30 R^2 = 0.77278 Chi^2/DoF = 14.56994 a 20.9788 ±1.73531 R^2 = 0.84231 b -0.19328 ±0.05202

a 22.75079 ±2.07217 b -0.28274 ±0.08227

20 20 20

10 10 10 Event mean concentration [mg/kg]

Event mean concentration [mg/kg] Event mean concentration [mg/kg] 0 0 0 0 1 2 3 4 5 0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8 3 3 Q tot [Mm3] Q tot [Mm ] Q tot [Mm ] • EMC for Copper and PAH show higher solid contamination in small events (Cu > PAH) 11 • Small events have higher contributions of first-flushes vs catchment erosion SPATIAL VARIABILITY Limited outreach of PAH pollution

200 40

Livange 150 30 Hesperange 35 Livange 200 Pfaffenthal Hesperange Pfaffenthal 30

150 25

100 20 20

100

16 PAH [mg/kg]

 15

PAH 16solidPAH [mg/kg]  EMC Cu solid [mg/kg] 10

EMC

EMC CuEMCsolid [mg/kg] 50 50 10 EMC 5

0 0

Event 1 Event 2 Event3 Event 6 Event 7 Event 8 Event 9 Event 1 Event 2 Event3 Event 6 Event 7 Event 8 Event 9 Event 10 Event 11 Event 10 0 0 Event 4+5 Event 4+5

Livange

Livange

Pfaffenthal

Pfaffenthal

Hesperange

Hesperange • PAH pollution and Copper pollution behave differently in longitudinal profile • Sources seem to be diverse and variably mobilisable 12 MASS FLOW COMPONENTS Deconvoluting flood waves Livange Wave 2 Real load Cu 9.07 kg FF load Cu 2.26 kg (28%) 7000 BG load Cu 4.27 kg (52 %) 200 FF2 load Cu 1.63 kg (20%) 6000 Sum load Cu 8.16 kg (fit 0.9) Discharge 150 5000

Particle solid concentrations /15 min]/15 Livange Wave 2 3000 3 4000 [mg/kg]: FF: 122 Real load ssp 129.90 t 100 BG: 44 7000 FF load ssp 18.45 t (15%) 3000 FF2: 161 BG load ssp 97.14 t (77%) 2500 2000 Cumin]load [g/15 6000 FF2 load ssp 10.10 t (8%) Discharge[m 50

Sum load ssp 127.7 t (fit 0.97) 1000 2000 5000 Discharge 0 0 0 50 100 150 200 250

/15 min]/15 3 4000 1500 time step [15 min] Livange Wave 2 Real load PAH 3.36 kg 3000 FF load PAH 0.29 kg (10%) 1000 7000 BG load PAH 2.44 kg (81 %) FF2 load PAH 0.27 kg (9%) 2000 Susp. M. [kg/15 min][kg/15 M. Susp. 60 Discharge[m 6000 Sum load PAH 3.00 kg (fit 0.89) 500 Discharge 1000 5000 Particle solid concentrations

/15 min]/15 40 3 4000 [mg/kg]: 0 0 FF: 16 0 50 100 150 200 250 BG: 25 3000 FF2: 27 time steps [15 min] 20

2000 min]loadPAH [g/15 • First we identify and fit SPM loads (turbidity signal) Discharge[m 1000

• Then we allocate a pollution to each SPM source 0 0 0 50 100 150 200 250 13 time step [15 min] MASS FLOW COMPONENTS Deconvoluting flood waves

Pfaffenthal wave 2 120 12000 Discharge Event First FF BG Real load PAH 6.85 kg load Flush conc. conc. 100 10000 FF load PAH 1.49 kg (21%) PAH load [mg/kg] [mg/kg] BG load PAH 4.38 kg (63%) [kg] fraction FF2 load PAH 1.14 kg (16%) 80 [%] 8000 Sum load PAH 7.01 kg (fit 1.02)

/15 min]/15 Livange 3.36 29 % 16 (27) 25 3 60 6000

Hesperange 3.61 22 % 33 (29) 18 40 4000 Solid concentration PAH:

Discharge[m FF: 48 mg/kg Pfaffenthal 6.85 37% 48 (13) 21 PAH solid[mg/kg] conc. 2000 BG: 21 mg/kg 20 FF2: 13 mg/kg

0 0 0 50 100 150 200 250 time steps [15 min]

• First flush contributions can be below background contamination • Contamination levels of first flush vs. background are variable downstream

14 LARGE SCALE PICTURE Catchment properties and PAH pollution

• Country wide sampling with 20 sediment nets at low-flow Sum 16 PAH [mg/kg]

• Large variety of catchment 15 properties (land use)

10

16 PAH [mg/kg] 16PAH 

5

0

Syr-Mertert Chiers- Attert-Bissen Alzette-Roeser Alzette-Essen Alzette-Beggen Eisch-HunneburSure-Bigonville Wiltz-KautenbachSure-Wasserbillig Alzette-Schifflange Mamer-Schoenfels Clerve-KautenbachErnz-blanche-Reisd

15 LOW FLOW PATTERN The random nature of PAH sources

Y = A + B * X Lead [mg/kg] 20 Sum PAH 16 [mg/kg] 600 Linear Fit of DATA1_D Parameter Value Error ------Linear Fit of DATA1_I A 16.09032 6.36148 Attert-Bissen B 1.78357 0.35721 400 ------Alzette-Beggen Ernz-blanche-Reisd R SD N P 15 ------Alzette-Roeser 200 0.83298 14.30142 13 4.0696E-4 Wiltz-Kautenbach 100 ------Clerve-Kautenbach Alzette-Beggen Attert-Bissen 10 Alzette-Essen 80 Chiers-Athus Alzette-Roeser Mamer-Schoenfels Chiers-Athus Sure-WasserbilligSyr-Mertert Alzette-Schifflange 60 Alzette-Schifflange Sure-BigonvilleEisch-Hunnebur 5 40 Sure-Wasserbillig Alzette-Essen

Ernz-blanche-ReisdWiltz-Kautenbach

Susp. matter content [mg/kg] Syr-Mertert Eisch-Hunnebur 20 Sure-BigonvilleMamer-Schoenfels Susp. matter content [mg/kg] Clerve-Kautenbach 0 0 0 10 20 30 40 50 0 10 20 30 40 Impermeable surface [%] Impermeable surface [%]

• Metals correlate well with impermeable surfaces while PAH do not at all • Alluvial contaminated sites as probable sources for PAH

16 TRACKING THE SOURCES How small scale is the problem?

• Combined passive sampler campaign (SPM nets & Empore disks for dissolved fraction) • Longitudinal stretch of 4 km length with 3 monitored sites • Different pollution sources • Urban + historical background (Livange) • Gas station (Berchem) • WWTP with known PAH pollution (Roeser)

17 SPATIAL DISCRIMINATION Sources reveal under low flow

Suspended matter 16 PAH [mg/kg] 6 30 Empore disk TWA [ng/L] 6 1500 Livange Discharge [m3/s] Discharge [m3/s] Berchem Turbidity [NTU] 5 25 Roeser Turbidity [NTU] 5 Livange Berchem 4 20 Roeser 4 1000

/s]

3

3 15 3

2 10 16PAH [mg/kg] 2 500 Discharge[m3/s]

Discharge[m



16 PAH dissolved16PAH [ng/L]  1 5 1

0 0 0 0 25/06/2014 09/07/2014 23/07/2014 25/06/2014 09/07/2014 23/07/2014

• Differences in SPM contamination can only be observed prior to the floodwave • Empore disk TWA show highest input by dissolved PAH downstream of the fresh

18 sources (WWTP, Gas station) FAR FROM EQUILIBRIUM

Apparent log Koc higher than literature values

10 Livange Fluoranthene Berchem Roeser 9

8

7

[L/kg OC] [L/kg

OC 6

5

LogK

4

3

2 lit value 30.6. 07.07. 16.07. 31.07.

• Apparent log Koc calculated with SPM and Empore disks are 3 orders of magnitude higher than expected from literature regressions

19 • Differences between the 3 sites are < 1 log unit and largest under low-flow conditions SOURCE DISCRIMINATION

The potential of log Koc to reveal aged contamination

40000 Alzette Longitudinal Profile8.4 Alzette Longitudinal Profile 35000 PAH Sum SPMf 8.2 30000 (ng/g) Log KOC (l/kg OC) 8.0 25000 20000 7.8 15000 7.6 10000 7.4 5000 7.2 0 7.0

• The south of Luxembourg is a historical steel working area

• SPM contamination and log Koc are decreasing downstream of the source

20 IMPLICATIONS FOR MOINTORING

Relevance of log Koc for whole water extraction?

1.0 Fluoranthene 1000 C 2 mg/kg s log K field 8.5 0.8 oc f 0.06 log Koc =4 oc log Koc =6 100 log Koc =8 0.6 total extractable whole water sample particulate conc. 10 dissolved concentration 0.4 AA-EQS water MAC-EQS water predicted dissolved concentration Fractioninwater[-] predicted whole water concentration

0.2 1 Fluoranthene measured Fluorantheneconcentration [ng/l] 0.0 0 50 100 150 200 0.1 Suspended matter conc. [mg/L] 0 50 100 150 200 250 300 350 400 450 500 Suspended matter conc. [ mg/L]

• With field Koc up to 3 orders higher than expected suspended matter concentration impacts EQS-evaluation heavily 21 • Are these EQS relevant under these conditions (bioavailability)? SUMMARY & CONCLUSIONS

• PAH EQS exceedences are one of the main reasons of chemical status failure • Several investigations with different approaches suggest that urban runoff is not the main source of PAH in catchments • Instead, suspended sediment profiles under low-flow suggest very localized contaminations • The outreach of these pollutions is very limited in longitudinal profiles (-> erratic conclusions on upstream situation)

• Apparent log Koc of suspended sediments are 3 orders of magnitude higher than literature values (implications for whole water sampling, SPM impact)

• Log Koc have the potential to discern fresh from old PAH pollution sources • Longitudinal profiles at low flow with combined SPM and Empore disk sampling can be used to pinpoint pollution sources

22