Source apportionment for Thessaloniki, Greece a PMF approach and a CAMx approach

1
Source apportionment for Thessaloniki, Greecea
PMF approach and a CAMx approach
APICE - Final Conference, 8th November 2012,
Venice
Aristotle University of Thessaloniki,
Greece Scientific Responsible Prof. Melas
D. Poupkou A., Liora N., Karagiannidis A.,
Markakis K., Giannaros T.
University of Western Macedonia,
Greece Scientific Responsible Prof. Bartzis
J. D. Saraga, E. Tolis, K. Filiou
2
Presentation overview

• Thessalonikis monitoring campaign
• PMF results (UOWM)
• CAMx results (AUTH)
• Discussion
• - Main conclusions
• - Mitigation Measures in the Future
• - Future Emissions

3
sampling sites
Thessaloniki city
PM2.5 sampling at two sites in Thessaloniki
CITY CENTER (TOWN HALL)
PORT
4
PORT
5
CITY CENTER (TOWN HALL)
6
Monitoring campaign schedule
1st period 14/6/2011-21/12/2011
2nd period 14/2/2012-22/5/2012
322 samples in total
7
24-hour PM2.5 samples collected on quartz fiber
filters using low volume samplers
TRACE ELEMENTS
PAH
IONIC SPECIES
OC/EC
8
Source apportionment study for Thessaloniki
PMF model application
results from 1st period
9
Source apportionment study for Thessaloniki
PMF model application
Positive Matrix Factorization (PMF) model a
widely used receptor model based on factor
analysis which provides a flexible modeling
approach using a set of data at a receptor
site, to indentify the unknown sources and
estimate their contribution
2 Selected sites port city center (Town hall)
period selected days from June-July-August
November-December 2011 PMF model analysis for
one- year data (June 2011 to May 2012) IN
PROGRESS
10
Source apportionment by PMF model
Chemical species data used for PMF analysis 7
(out of 27 measured) PAHs, Benzobfluoranthene,
Benzokfluoranthene, Benzoepyrene,
Benzoapyrene, Indenocdpyrene,
Dibenzoa,hanthracene, Benzoghiperylene.
Lighter PAHS were excluded from the analysis
because of their volatility 4 (out of 9
measured) ions NO3-, SO42-, Na, NH4 Cl- and
Na were excluded/ set as weak species
respectively because of their large analytical
errors. K, Mg, Ca2 ions were excluded as their
pair elemental species were used in the analysis
(to avoid double counting mass) 13 Metals Pb,
Ni, Cu, V, Mn, Cr, Zn, Mg, K, Ti, Fe, Ca,
Al Organic Elemental carbon OC, EC.

• experience obtained from the Intercomparison
  source apportionment analysis on Marseille data "
  was used

11
PMF results
Sampling site Thessalonikis port
? 5 groups of sources identified
mixed Vehicles/Anthropogenic Nitrates/Sulphates

combustion /central heating
mineral/ industrial
Mixed marine origin sea /fuel oil /ships
emissions
mixed Crustal/road dust/Incineration?
12
PMF results
Sampling site Thessalonikis port

• combustions/central heating source elevated
  during cold period

13
PMF results
Sampling site Thessalonikis city center (Town
Hall)
? 4 groups of sources identified
mixed nitrates /sulphates, anthropogenic Indust
rial
combustion /central heating
Possibly marine source included
mixed Crustal/ road dust
Mineral/Industrial
14
PMF results
Sampling site Thessalonikis city center (Town
Hall)

• combustions/central heating source elevated
  during cold period

• mixed source nitrates /sulphates,
  anthropogenic, Industrial/ ships emissions
  elevated during warm season

15
PMF results
Sampling site Thessalonikis port

• For the examined period
• crustal/road dust source combustions/central
  heating source similar contributions to PM2.5
  for both sites (appr.30)
• secondary aerosol/vehicles/anthropogenic source
  contribution elevated at city center area (51)
• mixed sea origin fuel combustion/ships
  emissions source contribute at the port area
  

Sampling site Thessalonikis city center (Town
Hall)
16
Source apportionment study for Thessaloniki
CAMx model application
17
Modeling System Set-up

• CAMx (version 5.3) simulates the Emission,
  Dispersion, Chemical reaction and Removal of
  pollutants in the troposphere on
  three-dimensional grid(s).
• Meteorology from the application of the
  meteorological model WRF (version 3.2.1)
• WRF-CAMx runs for the 2011 SUMMER PERIOD and the
  winter period 15 NOV TO 15 DEC 2011.

Chemical Boundary Conditions

• Gaseous and Particulate Chemical BCs for the
  Balkan domain taken from the global modeling
  system IFS-MOZART (within FP7 EU project MACC).

18
Emission Sources and PM Species Apportioned for
Thessaloniki

• The pollution transported to Thessaloniki from
  emission sources OUTSIDE the study domain will
  NOT be examined in this presentation.

19
Emission Inventory for the Maritime and Harbor
Activities
Emissions for the THESSALONIKI STUDY AREA
(Reference year 2010)
Emissions (tn/yr) CO NOx SOx NMVOCs NH3 PM10 PM2.5
Passenger ships 60.67 36.80 10.38 12.73 0.009 1.62 1.62
Cargo ships 881.02 7022.23 4537.84 113.92 0.904 220.79 220.79
Tugs 2.13 10.10 0.44 0.39 0.004 0.39 0.39
Inland waterways 26.78 72.87 - 9.58 0.011 7.59 7.15
Fishing Boats 315.75 3738.86 118.28 57.37 1.301 57.72 57.72
Harbor Operations (loading, unloading, pilling) - - - - - 37.2 5.64
Vehicles operating in the port - - - - - 0.181 0.043
Total 1286.35 10880.86 4666.93 193.99 2.23 325.56 293.36

• Most important activities in terms of emissions
  IN OR NEAR THE PORT 
• PM2.5 Ship Hotelling (Cargo ships).
• PM10 In-port Processes (loading, unloading and
  pilling of goods/materials).
• SOx, NOx Ship Maneuvering (Cargo ships) NOx
  ship hotelling emissions are comparable.
• CO, NMVOCs Ship Hotelling (Ferries) .

20
Source Apportionment for PM10 / PM2.5 Port
(Summer 2011)

• Road transport Highest contribution to PM10 and
  PM2.5 mean concentrations.
• Maritime/Harbor activities
• Moderate contribution to PM10 (14)
• Small contribution to PM2.5 (6)
• despite the emission contribution being 35 for
  PM10 and 15 for PM2.5.

21
Source Apportionment for PM10 / PM2.5 Port (15
Nov 15 Dec 2011)
20 for PM10 emissions
10 for PM2.5 emissions

• Road transport and Central heating Equal
  contribution to PM10 levels (30).
• Central heating Highest contribution to PM2.5
  levels.
• Maritime/Harbor activities Small contribution to
  PM10 and PM2.5.

22
Source Apportionment for PM10 / PM2.5 City Hall

• The contribution of the Maritime and Harbor
  activities is very small.

23
Maps of Contribution to Mean PM2.5 Levels
(Summer 2011)

• Maritime/ harbor contribution over sea
• ? PM10 up to 50 ? PM2.5 between 50 to
  75

24
Maps of Contribution to Mean PM2.5 Levels (15
Nov - 15 Dec 2011)

• Maritime/harbor contribution to PM10 and PM2.5
  over sea up to 20

25
With a View to the Future..

• Maritime/Harbor future emissions were calculated
  considering both the
• Port evolution (e.g. 6th pier extension,
  construction of a marina) (activity data provided
  by the Thessaloniki Port Authority SA)
• Future normative framework according to which the
  sulfur content in ship fuels should be reduced to
  0.50 m/m during cruising mode (sulfur content
  during maneuvering (in Greece) and hotelling mode
  is 0.1 m/m).

Difference between present time (2010) and
future time (2020) emissions
  CO NOx SO2 NMVOCs NH3 PM10 PM2.5
Passenger ships 35.22 25.93 -49.42 33.63 19.21 -6.06 -6.06
Cargo ships 29.37 40.73 -66.56 40.56 37.72 -63.92 -63.92
Inland Waterways 2168.66 2142.43 - 2152.89 2152.89 2163.64 2141.1
Fishing boats -35 -35 -35 -35 -35 -35 -35
Harbour operations (loading, unloading, pilling) - - - - - 78.01 77.82
Total emissions 58.48 28.80 -65.71 122.20 5.82 9.42 -1.23
26
Mitigation Measures in the Future

1. Cold ironing (zero ship hotelling emissions).
2. Use of chemical wetting agents to control the
   storage pile emissions (-90 of piling emissions).

Changes in maritime/harbor future time
emissions due to mitigation measures OVER THE
WHOLE STUDY DOMAIN
CO NOx SO2 NMVOCs NH3 PM10 PM2.5
-4.72 -1.46 -1.68 -4.64 0.01 -23.19 -9.00
Changes in maritime/harbor future time
emissions due to mitigation measures NEAR and IN
THE PORT
CO NOx SO2 NMVOCs NH3 PM10 PM2.5
-79.72 -45.60 -15.46 -81.51 0.00 -49.74 -52.28
27
Conclusions Contribution to PM levels

• PORT
• CAMx A) Moderate to small contribution to PM10
  (15) and PM2.5 (5) in summer, B) Small
  contribution to PM10 (8) and PM2.5 (3) in
  winter.
• PMF contribution to PM2.5 (22)
• CITY HALL
• PMF CAMx Very small contribution to PM2.5
  (lt5) both during summer and winter .
• Existing discrepancies because of
• Two different approaches that identify emission
  sources that partly match
• CAMx Maritime/Harbor sector Ships
  (un)Loading of goods Vehicles/Machineries
  (possible underestimation of emissions)
• PMF Marine sector Ships Sea Salt (possible
  overestimation due to the identification of
  contribution from sources other than ships)
• 
  OVER THE SEA (CAMx results)
• High contribution in summer Up to 50 for PM10
  and more than 50 (up to 75) for PM2.5.
• Moderate contribution in winter Up to 20 for
  PM10 and PM2.5.

28
Conclusions - Future Emissions

• In the year 2020, projected maritime/harbor
  emissions show
• Small increase in PM10 ( 10)
• Very small decrease in PM2.5 ( -1.2)
• Important decrease in SO2 ( -65)
• Moderate to important increases in CO, NOx,
  NMVOCs.
• Mitigation measures like Cold Ironing and the
  use of Chemical Wetting agents to control the
  storage pile emissions are expected to decrease
  by
• -50 PM10 and PM2.5 emissions in and near the
  port area
• -23 PM10 and -9 PM2.5 over the whole
  Thessaloniki study domain.