3. Overview of the Corinair 90 Process

3. OVERVIEW OF THE CORINAIR 90 PROCESS

CORINAIR 90 was extended to include countries outside the EU and now includes 30 countries (although not all of these have reported as yet). All of these countries agreed to contribute and have worked together to produce a single system for the whole of Europe. Definitions had to be agreed. Software was distributed and data collected by national experts.

This effort involved:

SNAP90 (Selected Nomenclature for Air Pollution 90), a source sector hierarchical nomenclature with 260 activities, three levels and 11 main sectors.
Extending the number of point sources.
Covered eight pollutants:
sulphur dioxide (SO₂)
oxides of nitrogen (NO_x)
non-methane volatile organic compounds (NMVOC)
ammonia (NH₃)
carbon monoxide (CO)
methane (CH₄)
nitrous oxide (N₂O)
carbon dioxide (CO₂)
Collaboration with UNECE which requires inventory information as part of the protocols of the Convention on Long Range Transboundary Air Pollution (LRTAP) and OECD (who were developing the IPCC methodology for greenhouse gas emissions). There has also been collaboration with the UNECE Task Force on Emissions Inventories that is producing the EMEP/CORINAIR guidebook on emission inventories based on the SNAP nomenclature.

When national CORINAIR 90 inventories have been compiled and checked, the data is then transferred to an ORACLE database held by the EEA, and the results of the European wide inventory are collated and distributed to users. Table 3 summarises the information which is contained in CORINAIR 90 and the requirements of some potential users. Table 4 is a summary of CORINAIR 90 data for Europe and an example of the more detailed data available from CORINAIR 90 is shown in Table 5.

Table 3 CORINAIR 90 Specifications and User Requirements

	CORINAIR 90	UNECE	EMEP	IPCC	EU Large Combustion Plant Directive	OSPAR & HELCOM	EU Greenhouse Gas Reporting	Other (e.g. Policy makers, NFP)
POLLUTANTS
SO ₂	·	·	·		·
NO _x	·	·	·	·	·
CO ₂	·	·		·			·
CH ₄	·	·	·	·
N ₂ 0	·			·
NMVOC	·	·	·	·
CO	·	·	·	·
NH ₃	·	·	·
OTHER				· HFCs · CF ₄ · C ₂ F ₆ · SF ₆		Heavy metals Persistent organo- chlorides	Other gases required by the IPCC but data requested not expected.	· Particulates · Specific VOCs
SOURCE CATEGORIES
CORINAIR SNAP LEVEL 1 - 11 GROUPS	·	·	·
CORINAIR SNAP LEVEL 2 - 57 CATEGORIES	·
CORINAIR SNAP LEVEL 3 - ABOUT 240 CATEGORIES ^,	·
OTHER	· Point Sources			· 6 groups split into 71 categories · Sinks (partial)	· Existing plant ³ 300 MW · Existing plant 50 to 300 MW · New plant ³ 50 MW		IPCC Source Categories	· DGXI VOC directive: to be specified · Policy makers: ISIC and socio -economic categories
SPATIAL RESOLUTION
NUTS LEVEL 0 (NATIONAL)	·	·	·		·	·	·	·
NUTS LEVEL 3	·							·
OTHER	· Individual large point sources as defined by CORINAIR 90		· 50 x 50 km grid		· Individual large point sources as defined in directive	· Not fully defined yet		· Individual point sources · Smaller grids · Urban inventories
TIME-SCALES
PROVISIONAL DATA							7 months
REVISED DATA		12 months	12 months	21 months	9 months		19 months

Table 4 CORINAIR 90 Summary for Europe

Table 5 Detailed CORINAIR 90 Data for France 1990 Emissions (as Mg except CO2 as Gg) for Industrial combustion plant and Processes with Combustion

SNAP	Process	SO ₂	NO _x	NMVOC	CH ₄	CO	CO ₂	N ₂ 0	NH ₃
030000	Industrial combustion plant and processes with combustion	514090	164965	7282	6623	598176	87391	2070	0
030100	Combustion in boilers, gas turbines and stationary engines	328113	66742	2398	2299	8328	44311	1570	0
030101	Combustion plants ³ 300 MW	113600	18556	603	565	2337	12653	458	0
030102	Combustion plants ³ 50 MW and < 300 MW	70365	17216	564	501	2028	11664	366	0
030103	Combustion plants < 50 MW	144116	29604	1200	1200	3871	19478	724	0
030104	Gas turbines	31	1366	33	33	91	517	23	0
030105	Stationary engines	IE	IE	IE	IE	IE	IE	IE	IE
030200	Process furnaces without contact ⁽¹⁾	63083	13959	594	586	5739	12247	283	0
030201	Refinery processes furnaces	49353	6246	205	202	1046	4854	190	0
030202	Coke oven furnaces	13196	7137	380	380	4282	1106	14	0
030203	Blast furnaces cowpers	0	454	0	0	394	6200	76	0
030204	Plaster furnaces	534	122	9	4	17	87	4	0
030300	Processes with contact ⁽²⁾	122895	84263	4290	3739	584109	30833	217	0
030301	Sinter plant	26389	20994	660	2423	549998	3080	178	0
030302	Reheating furnaces steel and iron	3569	2521	986	118	508	833	0	0
030303	Gray iron foundries	374	104	187	21	12456	83	0	0
030304	Primary lead production	21804	85	1	0	18	78	0	0
030305	Primary zinc production	13946	61	1	0	11	49	0	0
030306	Primary copper production	9	103	0	0	0	0	0	0
030307	Secondary lead production	3700	0	0	0	0	0	0	0
030308	Secondary zinc production	0	15	18	0	0	0	0	0
030309	Secondary copper production	27	5	80	0	10	0	0	0
030310	Secondary aluminium production	107	82	22	0	13	0	0	0
030311	Cement	16572	33666	1158	1158	9262	17946	0	0
030312	Lime (including iron and steel and paper pulp industries)	162	1750	14	0	174	2401	0	0
030313	Asphalt concrete plants	8640	576	227	0	1644	829	28	0
030314	Flat glass	22141	20946	0	0	288	3311	0	0
030318	Mineral wool (except binding)	NEG	NEG	NEG	NEG	NEG	NEG	NEG	NEG
030319	Bricks and tiles	3976	2565	282	0	8209	1657	0	0
030320	Fine ceramic materials	706	456	50	0	1458	294	0	0
030321	Paper-mill industry (drying process)	773	335	18	18	61	272	11	0
030322	Alumina production	0	0	0	0	0	0	0	0

Key: 0 = non-existing activity or no emission expected,
NEG = neglected
IE = included elsewhere)
(1) Processes where flames and/or combustion gases are not in contact with other products
(2)Processes where flames and/or combustion gases are in contact with other products

Note: CO₂ estimated as ‘"at source"

3.1 Achievements of CORINAIR 90.

While this paper will discuss in detail the problems of CORINAIR 90 it is important to remember that the CORINAIR 90 project has had many significant successes.

CORINAIR 90 is a major step forward in the compilation of a European inventory system that has achieved the highest level of completeness, consistency, comparability and transparency reached to date in such a wide international collaboration. While there are many different ways in which the individual country’s inventories differ the overall inventory is a major step forward in achieving its goals.

The collaboration, assisted by CITEPA, between EMEP and CORINAIR and other technical experts, has produced a system that covers 30 countries with a wide range of experience in the development of their national emission inventories. For some countries CORINAIR 90, was their first attempt at a national inventory while others already had a well developed national system.

CORINAIR 90 has resulted in a source classification, SNAP codes, that now has a wide acceptance in Europe. This is forming the basis of the joint EMEP/CORINAIR guidebook on emission inventories. Several countries wish to use CORINAIR 90 type inventories for data submission to UNECE (and EMEP) and IPCC (with the data conversion routines produced by CITEPA).

Unfortunately, the time-scale for completion of the CORINAIR 90 was not explicitly specified and adhered to as an important objective at the beginning of the project.

3.2 National Approaches to CORINAIR 90

An overview of possible national approaches to producing CORINAIR 90 is shown in Figure 1. Countries have taken different approaches to CORINAIR 90. The questionnaires (with 17 replies out of 31 participants) show that 4 countries used CORINAIR 90 to produce their national estimates, while 7 have ensured that their national estimates are consistent with CORINAIR 90. (Annex A describes the results of the questionnaires. Box 1 gives examples of the approaches taken in particular countries, and Annex B summarises the characteristics of each country’s CORINAIR 90 database).

Each country has reasons for adopting their individual approach and in the medium term the Emissions Inventory Topic Centre must either answer their concerns and needs or include them in any future methodology. A long term aim for the Topic Centre should be the adoption of a single methodology across Europe.

Figure 1 - Overview of Approaches to CORINAIR 90

Box 1 Examples of National Approaches to CORINAIR 90

In France, CORINAIR is used for the national estimates. All SNAP codes are treated, but sometimes there was considerable difficulty in doing this. The rubrics have been used either to describe technical differences (e.g. different cement processes) or to distinguish different economic sectors (e.g. industrial combustion is divided according to economic sector). Sometimes the softwares top-down approach has been used to spatially allocate data but at other times procedures outside the system have adopted a bottom-up approach. Some territorial unit specific emission factors have been used. Emissions measurements and other data (e.g. from mass balances) are included whenever they are available. A specific biogenic model is used for French biogenic emissions. In order to ensure that the transparency of CORINAIR is maintained, all activity rates and other data are provided in the inventory. This allows the recalculation of emissions factors and comparisons with other countries. Compiling CORINAIR is thus a mixture of a top-down and bottom-up approaches.

The Netherlands has a different system which measures or estimates emissions from many sources. The aim is to achieve a consensus system where the emissions are agreed by the individual companies and the contractors compiling the inventories (TNO). There is a specific problem in that the individual companies are allocated to economic categories according to their main income, which can be quite different from their emission activity. (For example, a manufacturing plant may be allocated to the trade sector.) They have entered the measured emissions into the CORINAIR 90 software with emission factors set to 1. Thus their contribution is not open and transparent as there is no way that their numbers can be compared with other countries contributions. As this combines lots of individual data about individual sources this is a bottom-up approach. The lack of transparency is also due to the CORINAIR 90 system expecting similar datasets to be collected for each country. As the original data collection and assumptions are open to limited inspection transparency could be achieved by looking at the original data.

Germany has two separate systems. Firstly the Lander estimated emission inventories for their areas. This is not done annually. However, due to legal restrictions separating the federal government and the regional Lander, these datasets are available to the federal authorities for specified uses only. The UBA (Berlin) has constructed an inventory for the whole of Germany and transferred this gridded dataset into the CORINAIR 90 system.

The UK has taken its existing methodology and entered the results into the CORINAIR 90 system. Thus the emission factors are taken from the national methodology. This has caused a few problems in areas where the sectoral splits do not exactly match. It does mean that the results are comparable and transparent as all the data used is entered into the system. As this works from national data this could be described as a top-down approach. It was important for the UK that there was only one estimate for the emissions rather than a set of estimates for different purposes (i.e. a national estimate, a CORINAIR 90 estimate, an EMEP estimate and a IPCC estimate). However this estimate could be revised in the future. In other words improvements in technical knowledge and the methodology are applied retrospectively. The UK believes the speedy production of time series and trend data is important.

There are a number of countries with developed, computerised systems where effort was expended transferring data into the CORINAIR 90 system. These include Norway, The Netherlands, UK and Sweden. As the CORINAIR 90 software is not the final product, that is the EEA ORACLE database in Copenhagen, these countries thought that transferring data was into CORINAIR 90 first was an extra, unnecessary effort. If countries were to transfer data directly to the Oracle database then it would be important that the quality of the data is still checked and that validation and verification procedures are carried out.

The Baltic states did not have experience in compiling emission inventories. They received considerable support from RISØ on behalf of CORINAIR 90 and were able to complete their initial datasets within six months. They found the provision of a complete system and guidance very useful.

In some other countries, no single organisation may be responsible for producing the national emissions inventory; experts may each contribute to a particular aspect, and may not use the same base data or the same assumptions.

3.3 Time Taken to Complete the Project.

CORINAIR 90 is only now starting to produce its results. A complete inventory will only be available five years after the end of 1990. CORINAIR 85 has only been completely reported in 1995. These long delays have obscured the many successes of the projects and have severely compromised the usefulness of the whole exercise. It is possible that when the data is finally published it is of limited use to policy makers. It is far too late to supply the data requirements listed in Section 5.3.

There are a number of reasons for this. They include:-

Time required. The effort needed ranged from 0.5 man-years to 5 man-years for different countries. This needs to be addressed. The Baltic States compiled provisional inventories within six months elapsed time. They had direct assistance and produced inventories with relatively few area activity rates.

Time waiting for other statistics. Many countries reported that waiting for data from others, either for the publication of official statistics or statistics from source sectors, caused delay to CORINAIR 90. This problem must be addressed, but this is best done on a country by country basis. In the case of official statistics, it may be possible to use pre-publication copies of the data, rather than waiting for them to be officially published. This is possible in the UK. In other countries where this is not possible estimates based on provisional international statistics are possible.

The effort required to learn to use the system. Two approaches to this problem can be considered. The software could be made simpler, or more direct assistance could be made available.

The low priority given to CORINAIR 90. This is a matter of policy in individual countries. Greater timeliness of the data would raise the profile of the work, and thus the pressure to meet the deadlines. Pressure from the EEA to collaborate would also help.

Shortage of funding and delayed contracts. In some countries a shortage of national or CEC funding (e.g. under PHARE), or a delay in receiving a contract from the national customer or the CEC, delayed completion of CORINAIR 90. As above a higher profile for CORINAIR work, and pressure from the EEA to collaborate might help. The switch to a rolling programme of work may also help.

The need to get internal agreement in an individual country to the data supplied. Again this is an policy consideration. In Belgium two inventories were compiled, one for each part, and this regional approach may help in other cases.

The amount and variety of the data requested. A lot of data was requested both sectoral and spatial. Despite requests from the EEA-TF little or no attempt was made to prioritise this and thus focus on the important parts. Similarly little attempt was made to collect fundamental data first (e.g. national totals) and the remaining data later.

Reluctance of experts to submit data that is not final or subject to revision. This is a particular problem with emission inventories as they are, by their nature, estimates. Thus many years can pass with experts ‘perfecting’ the estimates while users have to wait. Even then, these ‘perfected’ numbers are still estimates with, in some cases, large uncertainties. Another approach is to provide the best estimate that exists on a particular date. Provision can then be made for subsequent revision.

It is important to distinguish between technical problems that may delay the production of an inventory and structural or ‘political’ limitations. The former may include the late supply of data collected by others e.g. road transport statistics, while the latter may involve the need to get agreement from various official bodies about the data. Some of the ‘institutional’ problems which have occurred, e.g. delays in funding or contracts may be improved in Air Emissions 94 by the switch from a voluntary collaboration to a more structured framework of national focal points and national reference centres.