The resulting tables in the formats as described above are presented in Appendix 4. The first page of this appendix starts with an overview of the 40 (8 tables times 5 databases) tables. In this section the main outcomes of this overview are presented in an aggregated form and the most interesting features and differences are explained. It is explicitly not the aim of this section to present an analysis of what the role of wood energy is in Europe at present.

The overview is divided into (1) total wood energy use, (2) the wood energy consumption in the various sectors and (3) the different types of wood that are used.

Table 2 shows a selection of the main results of the tables in Appendix 4 in a very aggregated mode. One has to take care not to draw fast conclusions from the table itself, because the aggregated mode hides many structural differences between the various databases. In the text below the table the main figures will be elucidated by highlighting the main caveats per region. In the footnotes, general characteristics of the various databases are summarise.

• The large difference of about a factor 5 between FAO on one hand and UNECE/FAO and Eurostat on the other, is mainly caused by the fact that FAO includes only woodfuels that are directly removed from the forest, while the others include all wood and wood-derived fuels. This becomes very pronounced when looking at Table 3, where the total wood energy figures are disaggregated into individual countries. It can be seen there that especially in countries with much forest industry there is a large difference between FAO and the UNECE/Eurostat figures.²⁶
• In 1990 LBL and IEA comprise roughly half of the EU countries, but do include the largest ones (except France in IEA and Austria in LBL).
• The aggregated figures of Eurostat and UNECE are very close for 1990, although these institutes did clearly no make any direct use of each others figures. Table 3, however, shows that there are still considerable differences if the databases are compared country by country. The general trend is that Eurostat figures are higher for the Southern European countries and Denmark, lower for the new member states and quite comparable for the other (7) countries. The two differences appear to outweigh each other for a large extend in the total figure. A possible explanation for the higher Eurostat figures could be the inclusion of non-woody biomass (e.g. straw in Denmark and other agricultural residues in South-EU).

• In comparing the databases for the OECD-non-European countries, one has to realise that the countries included in this category differ largely. Therefore, also in this case a disaggregation into individual countries is presented (Table 4).
• The reason that the share in total removals is considerably higher for LBL compared with IEA (Table 2), despite of its lower absolute amount of energy, is that Canada is included in IEA and not in LBL. In the IEA database, Canada has a relatively large amount of total removals but a small share of wood energy use from these removals (about 20%) which brings the average for the region down.
• The figures for this region are clearly dominated by the USA figures. The estimates for the USA for LBL and IEA appear to coincide quite well. A reason for this could be that both are based on official national statistics. The UNECE figure for 1980 was 3,727 which is significantly higher. According to this database wood energy represents 4.9% of the total energy consumption and 88% of the total removals in 1980. The difference between FAO and the others is smaller here as compared with EU countries; the ration here is about 3.
• The presented growth figures are negligible for this region.

• Again, disaggregation into country levels is needed to make good comparison for the absolute figures for the various databases. This is presented in Table 5. It can be seen that FAO, UNECE and IEA included largely the same countries. Note that the LBL database only includes Norway.
• FAO figures are much closer to UNECE here than compared with the other two regions. One explanation for this is that also UNECE used much estimations for this regions, the fuelwood part of which was based on the FAO figures. Thus, partly, this is a resemblance by definition. The countries from which relatively complete replies were received by the UNECE questionnaires (Czech Republic, Norway, Poland, Switzerland) show much higher values than FAO (varying between a factor 2 to 6).
• Both UNECE and FAO show a slow decrease in the wood energy use that is included in their database.

Table 5: Wood energy in the various databases for European-non-EU countries in 1990 [PJ]

Disaggregation into the different wood-energy consuming sectors has been undertaken by UNECE, Eurostat and LBL. Figure 1 gives an aggregated overview of the results. All data are for 1990, except for the OECD-non-Europe data, which are for 1980.

The regions presented in this figure with the largest country gaps are : OECD-non-Europe for UNECE (only USA included), EU for LBL (only 6 countries, including Sweden, Finland and France), OECD-non-Europe for LBL (only USA and Japan) and Europe-non-EU for LBL (only Norway!).

In all three databases household consumption in the EU appears to be about 60% of the total²⁷.

The transformation sector for LBL is zero in all cases because it has simply not been included in their database (w.r.t. wood at least), since their transformation sector data do not rely on national statistics, but on IEA figures, which show no sectoral disaggregation at all for wood.

Figure 2 shows which types of wood are used for wood energy in the three regions according to the databases of UNECE and IEA. Again the OECD-non-European data for UNECE only comprise the USA and are for 1980 while all others are for 1990. Remarkable is the extremely high share of black liquor in the USA for the UNECE data. Although it is not clear whether their 1990 study [UNECE, 1990] uses the same method as the ETTS V study, part of the explanation could be in the method of converting the energy unit of black liquor into wood volume equivalents. In the EU, fuelwood directly from forests, appears to comprise about 40% in the two databases. According to UNECE, for European-non-EU countries this percentage is significantly higher (>60%).

In order to construct the analytical part of this report, a best estimate is made for wood energy use in EU, OECD-non-European countries and Europe-non-EU.

The two most reliable, complete and disaggregated databases for the EU are the UNECE and Eurostat database. It is preferred however to use one single set of data for the analysis chapter. Therefore the strong and weaker points of the two databases are weighed against each other:

Eurostat

Strong points:

• All data are compiled by national expertise centres. No estimates by Eurostat itself are included.
• The national centres of expertise were able to spend time on collecting data for their background reports, because they were paid for it under contract by Eurostat.
• The sectoral consumption data of Eurostat fit within the format of the energy balances as applied by IEA, Eurostat and UNECE.
• The database is based on the sectoral consumption. Especially household wood energy consumption data are of high quality, because of the specific surveys that were undertaken for this purpose.
• Eurostat database is improved and updated every year. Timeseries are produced.

Weaker points of the Eurostat database are:

• The database is very aggregated with respect to the various supply sources. There is only one category of wood/wood waste/other solid waste which for some EU countries is likely to include a significant amount of agricultural residues.²⁸
• The quality of the data for the new member states is still lower. For Sweden and Austria only total biomass/waste estimates have been given (also including MSW, anaerobic fermentation and biofuel production). These new member states are very important with respect to wood energy. It is clear that this lack in the database will improve in the future.
• No data prior to 1989 are available.

The strong points of the UNECE/FAO database are:

• The database is explicitly focused on wood energy only.
• The database is based on a high level of disaggregation on supply sources; various different types of wood are distinguished.
• Data are available back to 1980.
• All large wood energy consuming EU countries are included within the nine countries of which complete questionnaire replies were received.
• There is already about 20 years of experience within UNECE in collecting and interpreting wood energy data.

Weaker points:

• There are no timeseries. The gap between the data as presented in ETTS V is ten years, which is very large. It is a pity that the 1985 information from either the questionnaire or the 1990 study [UNECE, 1990] were not included in the analysis.
• The definitions of the sectors are not completely in accordance with the ISIC standards and do as such not fit in the standard energy balances.
• Sectoral consumption data are only available for 6 EU member states.
• The secretariat had to make estimations for 6 countries. The estimations by the secretariat are not always consistent with those made in the original report by Morin.
• -  No timeseries are available, because it is believed that the quality of available data at the national level does not allow for such a detailed assessment.
• The conversion of black liquor is believed to lead to an overestimation of about 25%.

Although the choice for one of the two would provide the most consistent set of data it is decided to use both databases:

• The overall wood energy consumption (for 1980 and 1990) and the disaggregation into the various types of fuels will be based on the UNECE data.
• The 1985 UNECE questionnaire data are added.
• For all categories of wood energy, the number of the preceding year is repeated, if for a certain country data are not available for all three years (but at least for one year). If there is no data in a preceding year or if only the 1980 figure is known , data of the following years are repeated.²⁹
• If for the categories "direct woodfuels", "primary residues" and "black liquor" for a certain country no data were available at all, they were estimated. The UNECE estimations in their final report [UNECE, 1996] are adapted so that they are in consistence with the estimations in the original UNECE background report made by Morin [Morin, 1995]. This implicates the following:

- The "direct woodfuel" figure is estimated by the FAO fuelwood figure [FAO, 1996].

- The "primary residues" figure is estimated on the basis of the relation that can be observed in look-alike replying countries³⁰ between their primary residue figure and their figure for veneer and sawnwood production from FAO [FAO, 1996].

- The black liquor figures have been estimated on the basis of the FAO chemical pulp figures [FAO, 1996], according to the formula used by Morin.³¹

- If other figures than direct woodfuel, primary residues and black liquor were not available for a certain country, it has been estimated as 0.

• The sectoral consumption (not in absolute values but in terms of percentages) is based on the Eurostat database.³²

In this way it is believed that the strong points of both databases are combined, without doing too much concessions to the consistency of the set of data.

The resulting set of data will be further presented in the analysis part of this report.

There is a clear lack of reliable data on wood energy from this region. UNECE and Eurostat, which were the most reliable databases for the EU, did not include this region, with the exception of the 1980 USA figures in the UNECE database.³³ The FAO figures are believed to present huge under-estimations if one considers total woodfuel use. The two databases left are LBL and IEA. The LBL database is clearly based directly on national statistics and is therefore believed to be the most reliable of the two. Therefore we will use the LBL figures for the USA and Japan, and the IEA figures for Australia, Canada and New-Zealand. Because of the low reliability, no best-estimate for this region will be presented on disaggregation to sectoral consumption.

Here the UNECE does present data for all countries. However only figures for four countries of the total of 22 were based on questionnaires. Figures for the other countries were estimated as described in chapter B.3. Fuelwood was estimated on the basis of the FAO figures and primary residues on the basis of comparing their production of veneer- and sawnwood with countries that did reply.

However, since the UNECE appears to be the least unreliable, and the most complete, we will use this database (with caution) in the analytical part of this report. Again, no disaggregation into the various consuming sectors will be made here, because of the low amount and low quality of data.

The recommendations will be split in two parts. One are general recommendations to the whole "wood energy statistical society" and the other are specific recommendations to FAO with respect to the establishment of a world wide database on wood energy.

• Although it is very clear that communication between the various statistical institutes that deal with wood energy has rapidly increased the last year, an explicit (ad-hoc) meeting of representatives of all relevant institutes (at least the ones mentioned in this report) could be very desirable. A possible initiator of such a meeting could be the already existing inter-secretarial Working Group on Forest Statistics. Though, it is desirable that the focus will be mainly on wood energy.

In the following some points of attention will be mentioned that could be dealt with during such a meeting:

• More integral co-operation between all institutes (so not only bilateral agreements) could lead to:

- increased consistency between all wood energy databases

- more efficient use of the limited resources in all institutes

- higher quality databases by combining expertise on both forestry and energy statistics

- better availability of country respondents by avoiding that information is asked more than one time.

• There should be more consistency in definitions with respect to both the fuel types and the consuming sectors. The highest level of disaggregation of fuel types should preferably be based on the inherent quality of the fuel (e.g. difference between woodfuels and wood-derived fuels, such as black liquor). Further disaggregation that fits within such a basic disaggregation by individual institutes according to specific wishes is then still possible. Disaggregation into consuming sector should fit within existing energy statistics and be according to the official ISIC categories.
• The treatment and conversion of units is an important point of attention. The way in which primary data are converted should be the same at all institutes. Conversion from volume to energy units needs some basic assumptions, to be agreed upon by all relevant statistical institutes. The main are the average dry matter density of the wood that is used and its average moisture content. On the basis of this, conversion factors can be calculated. It has appeared that presently there is no clearness on what kind of assumptions are appropriate. It should be realised that huge distortions can occur with this.³⁴ It may be necessary not to have overall conversion factors, but to differentiate for regions (e.g. Nordic countries, Central and Southern Europe) according to the types of trees that are used for fuelwood. In this case assumptions can be different for different regions, but the method for conversion should remain the same.
  The conversion method to express black liquor into wood volume equivalents should get special attention.
• The present national respondents should be made aware of available statistical data that has been prepared for their country on behalf of other statistical institutes. Activities that enhance this, are going on at IEA at the moment, where one explicitly points the national respondents (in case the reliability of data is suspected to be too low) at the existence of the Eurostat country studies. Activities like this could be extended by Eurostat and IEA with respect to the figures that have been compiled by the respondents of the ETTS V questionnaire of UNECE (and vice versa, of course).
• A discussion could be undertaken on the most suitable ways of data collection on the item of wood energy. Experiences at FAO and IEA have learnt that including wood energy within a huge questionnaires on other items (respectively the forestry and coal questionnaire) may lead to neglecting this item. The results of the databases of Eurostat and UNECE, however, states that energy from wood is a very, but complicated energy source and therefore deserves more attention.

• The present way in which wood energy is dealt with in forestry statistics seems quite undesirable. At least, the definition of fuelwood within the present forest product yearbook should be more clear. If one aims at including direct removals of wood for energy in order to complete the balance it should be very clear from the name and the definition of this wood that this is not representative for the total use of wood for energy. The present name of "fuelwood" and the lack of a suitable definition in the yearbook does not meet this requirement. An alternative name and definition could be the one that has been used in this report: "direct forest-woodfuel" (see Appendix 3).
• If FAO aims at having a complete database on energy from wood, the present statistical database should be extended. Of course, this should always be in line (w.r.t. definitions and units) with existing databases and it should be avoided that questionnaires ask for information which is already asked for at other places.
• One should consider whether it is necessary to collect additional data at member countries or to make use (possibly in co-operation) of already existing sources. It can be questioned whether inclusion into the present forestry statistics will guarantee sufficient attention to this complicated matter of energy from wood.
• Inclusion of a complete database on wood energy may not fit within the present structure of the FAO forestry statistics. The main balance here is in the total removals, while wood energy is created to a large extend downstream, from sources that have been put in the production statistics under other categories (like pulp- or sawnwood). Of course, double counting should be avoided. Either the indirect wood energy use should be a separate item outside the removal balance or the balance should be constructed as a complete material flow balance. The latter has the main problem of data-availability and the time-aspect. This time aspect (wood can be available for energy as a recycled product many years after its harvest) is a major complication as compared with the balance within energy statistics.

It is up to the FAO to decide to what level of detail data on energy from indirect woodfuels and wood derived fuels have to be integrated within the present forestry statistics.

B.7. References

Broek, 1996: Review on data on woodfuel use, production and trade in European countries: a preparatory document for the regional "Wood energy today for tomorrow" study on European countries., Department of Science, Technology and Society, Utrecht University, Utrecht.

Chipeta, M., 1996: Personal communication, FAO, Rome, 1996.

Denman, J., 1996: Personal communication, IEA, Paris, 1996.

Eurostat, 1993: Energy consumption in households, , Luxembourg.

Eurostat, 1995: Renewable energy sources statistics: 1989-91, , Luxembourg.

Eurostat, 1996: Renewable energy sources statistics: 1989-1994, , Luxembourg.

FAO, 1996: FAO forest product yearbook, , Rome.

IEA, 1996a: Energy balances of OECD countries: 1993-1994, IEA, Paris.

IEA, 1996b: Energy statistics of OECD countries: 1993-1994, IEA, Paris.

IEA and ENECE, 1995: Coal (solid fuels, wastes and manufactured gases): annual questionnaire: 1993 and 1994, , Paris.

Morin, 1995: Energetic use of wood (in French), UNECE, Paris.

Padovani, 1996: Personal communication, FAO, Rome, 1996.

Prins, C., 1996: Personal communication, UNECE, Geneva, 1996.

Roubanis, N., 1996: Personal communication, Eurostat, Luxembourg, 1996.

Savage, E., 1996: Personal communication, IEA, Paris, 1996.

Schipper, L., 1996: Personal communication, Lawrence Berkeley Laboratory, Berkeley, 1996.

UNECE and FAO, 1990: Survey of medium-trends for wood raw material, notably pulpwood, and wood for energy, Timber bulletin. XLII(2).

UNECE and FAO, 1996: European timber trends and prospects: into the 21st century, Report no. Geneva timber and forest study papers, Geneva.