PARTICULATE MATTER EMISSIONS FROM RESIDENTIAL WOOD COMBUSTION Revised estimates and an evaluation | Hugo Denier van der Gon CO-AUTHORS: Antoon Visschedijk, Christer Johansson, Robert Bergström, David Simpson, Christos Fountoukis, Spyros Pandis

CONTENTS… A STORY IN CHRONOLOGICAL ORDER The FP6 EUCAARI EC and OC inventory Application – model results vs observed Zooming in on the blooper….Wood combustion An alternative inventory for wood combustion Repeat application… Conclusions Particulate matter emissions from residential wood combustion

ACCENT+ Urbino, 15/9/2011Hugo Denier van der Gon - wood combustion PM 3 ANTHROPOGENIC CARBONACEOUS PM EMISSIONS IN EUROPE 2005; TNO EUCAARI EC/OC INVENTORY (KULMALA, 2011) Residential combustion dominates EC & OC calculated as a fraction of PM2.5

ACCENT+ Urbino, 15/9/2011Hugo Denier van der Gon - wood combustion PM 4 Note: wood combustion & transport

ACCENT+ Urbino, 15/9/2011Hugo Denier van der Gon - wood combustion PM 5 PMCAMX EVALUATION (WINTER) OA Finokalia Similar performance in Cabauw, Mace Head, Puy de Dome Pretty good

ACCENT+ Urbino, 15/9/2011Hugo Denier van der Gon - wood combustion PM 6 PMCAMX EVALUATION (WINTER) Similar performance in Payerne, Melpitz OA Vavihill A lot seems to be missing!!! But many improvements made and bulk of residential PM2.5 emissions allocated to OC….a puzzle…

Particulate matter emissions from residential wood combustion EMISSIONS OF PM10 PER UNIT OF WOOD COMBUSTED IN THE RESIDENTIAL SECTOR IN 2009 (G/GJ) Why rethink PM from wood combustion?

RETHINK THE RWC INVENTORY A first comparison/ analysis showed incompatibility of different country approaches (not “right or wrong” ) Start partly from scratch Don’t use national reported data Do not define EC and OC as a fraction of PM10 or PM2.5 that should “fit” in the official reported data Make a selection of emission factors and apply in a consistent way for all countries “independent bottom-up” 8 | Particulate matter emissions from residential wood combustion

FILTERABLE PM AND CONDENSABLE PM the US EPA defines particulate matter (PM) as consisting of a filterable fraction (FPM) and a condensable fraction (CPM). Particulate matter emissions from residential wood combustion Filterable PM is directly emitted: Solid or liquid Captured on filter PM 10 or PM 2.5 Condensable PM is in vapor: Reacts upon cooling and dilution Forms solid or liquid particle Always PM 2.5 or less where should the PM mass be that forms almost instantaneously?

PM EMISSION FACTORS FOR RWC EXAMPLE: CONVENTIONAL WOODSTOVE Source: Nussbaumer et al. (2008)

Source sectors EC_2.5OC_2.5 CARBONACEOUS AEROSOL < 2.5 UM IN UNECE-EUROPE FOR AFTER REVISING RESIDENTIAL WOOD COMBUSTION ONLY - Particulate matter emissions from residential wood combustion + 30 kt EC kt OC Limited impact on EC, major change in OC Changes in individual countries differ from European average

WOOD USE IN NORWAY < SWEDEN Wood use in Sweden larger but different appliance Countries use their own methods to calculate PM emissions Particulate matter emissions from residential wood combustion

NORWAY & SWEDEN TOTAL RWC EMISSIONS YEAR 2005 Particulate matter emissions from residential wood combustion

Comparison of model calculated OC from wood burning to source-apportionment data from measurement campaigns during winter in Norway and Sweden (SORGA, March 2007, Yttri et al.,2011) (GÖTE, 11 Feb - 4 Mar 2005, Szidat et al., 2009). Unit: µg(C) m −3 While we adjusted emissions in opposite directions for NOR & SWE, the model prediction improves at all sites!

Particulate matter emissions from residential wood combustion BIOMASS BURNING EC (EC BB ) FROM WINTER TIME SOURCE APPORTIONMENT STUDIES IN SCANDINAVIA COMPARED TO MODEL CALCULATIONS USING DIFFERENT EMISSION INVENTORIES. Bars represent 10th and 90 th percentiles of observed EC from biomass burning. Note the logarithmic scale on the y-axis. Source: Genberg et al., Atmos. Chem. Phys., 13, , 2013

RATIO RWC OC EMISSION BY COUNTRY (NEW/ORIGINAL ) Particulate matter emissions from residential wood combustion Red new inventory is higher; Green new inventory is lower

IMPACT OF CONSISTENT BOTTOM-UP CALCULATION FOR WOOD COMBUSTION EMISSION FOR SELECTED COUNTRIES Particulate matter emissions from residential wood combustion Indication of impact: 20% more PM2.5 emission over Europe

MEASURED AND MODELLED OC CONCENTRATIONS WITH THE EMEP MSC-W MODEL Particulate matter emissions from residential wood combustion

MEASURED AND MODELLED OC CONCENTRATIONS WITH THE EMEP MSC-W MODEL Particulate matter emissions from residential wood combustion

FINISHED? ….NO! ….SOME FURTHER REMARKS ON THE RWC EMISSIONS Wood use is notoriously uncertain; New update of wood use for year 2010 – = +28% more wood use than in our 2005 adjusted figure (used in this study) which was already + 20% above “official” (with large inter-country variation) Uncertainty ranges not properly covered: – quality of wood, stoves, burning practises, all greatly influence total (condensable) PM emissions. Distribution in time (within the year) and space (within the country) can significantly improve [and we know how] Close collaboration between Emissions – Measurement - AQ model communities can lead to much further improved understanding – including quantification of episodes and exceedances …..But over the past few years not able to find funding for such integration 21 | Particulate matter emissions from residential wood combustion

CONCLUSIONS Including condensable PM emissions for RWC greatly improved our understanding of measured OC (and thereby PM2.5) in ambient air at all locations. In line with US-EPA “the emission-based RWC contribution to ambient carbonaceous PM2.5 predicted by the model is approximately a factor of two lower than indicated by observations”. [Napelenok et al., 2014.] Measurement data are crucial for “proofing” but needs more than just PM2.5 We argue that condensable PM should be included in the PM inventories but it’s not “free” ; it will change estimated PM2.5 emissions Particulate matter emissions from residential wood combustion Our current primary PM emission inventories are “apples and oranges” and might well be a whole fruit basket…. Note: This revised TNO RWC inventory is NOT included in the TNO-MACC inventories

THANK YOU FOR YOUR ATTENTION Denier van der Gon et al., ACP, 2015

RELATED & SUPPORTING PUBLICATIONS Denier van der Gon, H. A. C., Bergström, R., Fountoukis, C., Johansson, C., Pandis, S. N., Simpson, D., and Visschedijk, A. J. H.: Particulate emissions from residential wood combustion in Europe – revised estimates and an evaluation, Atmos. Chem. Phys., 15, , doi: /acp , Simpson, David, and Hugo Denier van der Gon, “Problematic emissions - particles or gases?”, chapter 5 in: European Monitoring and Evaluation Programme (EMEP) Status report 2015, Fountoukis, C., T. Butler, M. G. Lawrence, H.A.C. Denier van der Gon, A. J. H. Visschedijk, P. Charalampidis, C. Pilinis, and S. N. Pandis, Impacts of controlling biomass burning emissions on wintertime carbonaceous aerosol in Europe, Atmospheric Environment, 87, p DOI: Genberg, J., Denier van der Gon, H. A. C., Simpson, D., Swietlicki, E., Areskoug, H., Beddows, D., Ceburnis, D., Fiebig, M., Hansson, H. C., Harrison, R. M., Jennings, S. G., Saarikoski, S., Spindler, G., Visschedijk, A. J. H., Wiedensohler, A., Yttri, K. E., and Bergström, R.: Light-absorbing carbon in Europe – measurement and modelling, with a focus on residential wood combustion emissions, Atmos. Chem. Phys., 13, , doi: /acp , Bergström, R., Denier van der Gon, H. A. C., Prévôt, A. S. H., Yttri, K. E., and Simpson, D.: Modelling of organic aerosols over Europe (2002–2007) using a volatility basis set (VBS) framework: application of different assumptions regarding the formation of secondary organic aerosol, Atmos. Chem. Phys., 12, , doi: /acp , | Particulate matter emissions from residential wood combustion

SAMPLING PROTOCOL AND MEASUREMENT TECHNIQUES STRONGLY IMPACT EMISSION FACTORS Industrial stack PM emissions are defined as particulates (aerosols) that are present in waste- or flue gas streams and not the condensable gases in these streams. Condensable particulate matter is by definition excluded Road transport exhaust PM emissions by definition capture most of the condensable PM because measurements include dilution and cooling to 51 o C. For residential combustion stoves various methods circulate which capture the condensable PM in various degrees (from 0-100%) How should AQ modellers know about this, when the inventory compilers don’t even know….. Particulate matter emissions from residential wood combustion Emission inventory compilers work source by source, compiling the best Emission factor by source but do not compare across sources