3:ANALYSIS OF AVAILABLE WOODFUEL STATISTICS AND DATA

3.1. Introduction

African countries rely on wood to meet energy requirements far more heavily than most other developing countries.

However, while woodfuel utilization has traditionally been a crucial issue as regards its interaction with environment, development and social welfare, there have been no serious efforts in any African country to include it definitively as a basic sector in planning processes.

Given this situation, the collection, compilation and analysis of biofuel data, particularly woodfuel data, never constituted a priority for these countries, and past initiatives (surveys, studies) were more project-based and tended to be limited in scope from the following points of view:

· geographically (e.g. : limited area, rural or urban only);

· sectorally (e.g. : only a single sector such as residential, or a single end-use);

· historically (no serious consistent historical data allowing comparability of data and trends analysis).

In order to achieve state-of-the-art available woodfuel information in Africa, FAO has recently launched a study aimed at compiling, comparing and assessing all existing international and national data, and deriving the best estimate. In addition to international consultations, extensive consultations with country experts on woodfuel matters are being launched in order to validate the best estimates and come forward with relevant comments and recommendations for the next phase, which will address the improvement of woodfuel information systems in Africa.

3.2. Methodology and approaches used

In order to meet the objectives of the current study, it was necessary to adopt a step by step approach :

Step 1: Consultation and assessment of the main available databases and documentation at international level (FAO, IEA, etc.). This step should have provided a rough idea of the main issues regarding the compilation of African woodfuel data at international level.

Step 2: Consultation and assessment of the main available national statistics at FAO, IEA and consultant libraries. This step is very important as it will help to provide a more accurate idea of existing data at national level, and to identify the major weaknesses.

Step 3: Comparison of national data with the data included in the international databases in order to define a preliminary idea of "best currently available estimates". While these estimates are based on existing materials, the consultant's experience in Africa can help fill in the gaps.

Step 4: Presentation of the selected consistent data as well as the best estimates using the Unified Wood Energy Terminology.

Step 5: Identification of ways to improve data quality, and defining a relevant approach for the collection, compilation and presentation of woodfuel data in Africa.

All data sources are individually presented using a unified format and articulated in the following major tables :

1. Presentation of a time series (ranging at least from 1980 to 1995, and even outside this range where data are available, on global woodfuel consumption (fuelwood and charcoal aggregated on the basis of primary wood) using to model ("Table A.x.y" ⁽⁶⁾):

· In terms of volume: (Table A.x.1 woodfuel consumption m³ )⁽⁷⁾
· In energy terms : (Table A.x.1: petajoules⁽⁸⁾)

2. Presentation of the contribution of woodfuel consumption to :

· National wood removals (Table A.x.2)
· National primary energy consumption (Table A.x.4)

3. Presentation of the supply side of woodfuel consumption : Table A.x.5 ( m³) and Table A.x.6 (PJ). The figures presented are :

· Direct forestry woodfuels
· Indirect woodfuels (industrial wood residues used as fuel, black liquor, etc.)
· Recovered woodfuels

4. Presentation of the demand side of woodfuel consumption : Table A.x.7 (m³) and Table A.x.8 (PJ). The different concerned sectors are :

· Households
· Industries
· Transformation
· Others

In addition, other tables were developed in order to record all original data collected. These tables are included in each individual appendix of the different data sources.
A specific effort was made in the case of developing countries, with one table used to present the figures recorded by the five data sources as well as the best estimates for each country, thereby enabling data consistency to be crosschecked.
The countries are grouped on a regional basis given regional similarities in interactions between the geographical location and socio-economic and ecological contexts, which are themselves closely linked to woodfuel practices. Eight groups of countries were identified :

· West Sahelian Africa
· East Sahelian Africa
· West Moist Africa
· Central Africa
· Tropical Southern Africa
· Insular East Africa
· North Africa
· Non Tropical Southern Africa

3.3. Description of the databases

Almost all the woodfuel data in Africa available at the time of this study were consulted and compiled. At first, compilation only concerned the two formal databases : the FAO database and the IEA database.

In addition, sources which could be more appropriately termed "statistics" rather than databases were also assessed. While these statistics presented a generally high degree of uncertainty, consistent statistics based on field surveys were also found and integrated into analysis in order to correct the other data or to fill in gaps for some countries. These data could also be used to complete the time series. Data assessed came from the following sources :
· ESMAP statistics
· ENDA/IEPE statistics
· Various national data
· Other international sources of woodfuel African statistics (African Development Bank, United Nations Statistics, World Resources Institute statistics, etc.)

3.3.1.The FAO database (FAO Forest Products Yearbook)

The data collecting approach

FAO publishes a statistical yearbook on forest products, including fuelwood and charcoal (production, imports and exports). The FAO database also presents aggregate figures of fuelwood and charcoal in terms of woodfuel equivalent (m³).

The data published in this yearbook are theoretically collected through official channels by sending a questionnaire to specialized departments in the countries. However, given the poor data quality at national level, particularly on woodfuels, and the low response rate from various countries, this surveying approach has proved unsuccessful.

Almost all woodfuels in Africa are collected, transformed and consumed through informal channels. Thus such traditional practices limit prospects for handling statistical data collecting systems. Moreover, difficulties in observing woodfuel behaviour trends can hamper national capacity to provide regular updated estimates. Therefore the approach of annual data collection and compilation through international surveys seems inappropriate and a new approach is called for.

In order to present historical series, FAO makes simple estimates based on per capita consumption and population growth. However, the quality of FAO data appear uncertain for the following reasons:

· The approach used to fill the gaps is not documented in the statistical publications for Africa (although it is indicated in the electronic database that all the data on woodfuel production and consumption originated from FAO calculations).

· The exact origin of the annual published data for each country (primary data, derived from the FAO calculation, etc.) is not specified.

· The reference data used to derive the time series are not specified.

· In some cases, the same level of per capita consumption is used for the whole time series.

· In other cases, the global consumption of a woodfuel item (e.g. fuelwood) remains unchanged for many years.

Therefore, FAO data should not be considered definitive, and are permanently subject to changes from one publication to another in line with new data acquired or with possible changes in the estimating approach.

Structure of the database

The FAO database presents historical series of woodfuel use under three items :

Fuelwood data (distinguishing Coniferous from Non Coniferous)

· Production (assimilated to consumption) of fuelwood in terms of 1000 m³;

· Imports of fuelwood in terms of 1000 m³

· Exports of fuelwood in terms of 1000 m³

Charcoal data

· Production of charcoal in terms of 1000 m³

· Consumption of charcoal (assimilated to production while in some cases no consistent relation are noted between production-imports-exports and consumption) in terms of 1000 m³

· Imports of charcoal in terms of 1000 m³

· Exports of charcoal in terms of 1000 m³

Aggregate woodfuel data

· Production of woodfuels (aggregate figures of fuelwood and charcoal) in terms of 1000 m³

· Consumption of woodfuels (assimilated to production) in terms of 1000 m³

· Imports of woodfuels in terms of 1000 m³

· Exports of woodfuels in terms of 1000 m³

Regarding woodfuel trade among African countries or with other non African countries, the FAO database is the only one which reports annual figures on imports and exports of fuelwood and charcoal. The data included on the database generally come from official country sources and can be considered as the formal declared figures, while figures on informal trade among countries are not documented anywhere.

In general, official trade in fuelwood and charcoal is insignificant compared to total African consumption. For example, charcoal imports totaled 57 000 tons for the whole of Africa in 1994 and 59 000 tons in 1995, while exports represented only 80 000 tons in 1994 and 78 000 in 1995.

Fuelwood imports amounted to 7 000 m³ for the whole of Africa in 1994 and 45 000 m³ in 1995, while exports represented only 276 m³ in 1994 and 6 000 in 1995.

Although trade volumes in Africa do not represent a great issue, FAO statistics provide no detail on the originating country of woodfuel imports for each country nor on the destination of exports. This makes it is difficult to check the consistency of data submitted to FAO, at least for African countries.

For example, gaps between imports and exports might be considered consistent since charcoal imports to Africa and exports from Africa can be envisaged within the economic framework of charcoal trade. Conversely, this might not be true for fuelwood, where the gaps between imports and exports could be explained by possible data inconsistency rather than potential trade with other continents.

Approach for using the data in the context of the study

According to FAO definitions of forest products, fuelwood normally includes raw wood from the trunk and branches of trees. However as information in Africa tends to originate from demand, the term fuelwood has come to include all sorts of wood from different tree sources, including fruit trees. Charcoal could be viewed in a similar way as, again, data are generally available from the consumption side.

On the other hand, FAO statistics distinguish between non-coniferous and coniferous woodfuels; however, this distinction was not considered relevant in this study as it does not represent a key issue for woodfuels.

FAO data calculate woodfuel consumption as follows :

· calculation of fuelwood consumption (1000 m³) = production + exports - imports;

· calculation of charcoal consumption (1000 m³) = production + exports - imports;

· calculation of aggregate woodfuel consumption (1000 m³) = consumption of fuelwood + 6 x consumption of charcoal⁽⁹⁾;

The results of data compilation were finally presented in the following uniform tables:

1. Main tables :

· In terms of volume (Table A.1.1 : woodfuel consumption : 1000 m³, and Table A.1.1 Per Capita woodfuel consumption : m³ /year).
· Aggregate woodfuel consumption as a percentage of total wood consumption (Table A.1.2 : woodfuel consumption as a percentage of total wood consumption).
· In energy terms (Table A.1.3 : Petajoules).
· Direct and indirect woodfuel consumption (Table A.1.5) in terms of m³ .
· Direct and indirect woodfuel consumption (Table A.1.6) in energy terms (PJ).
· Aggregate woodfuel consumption by sector (Table A.1.7) in terms of m³. However this table was not relevant for FAO since its database does not include any details on consumption share.
· Aggregate woodfuel consumption by sector (Table A.1.8) in energy terms. However, this table was not relevant for the same reason as above.

2. In addition, original tables were also included in the datafile :

· Total Roundwood consumption (Table A.1.1a ) in terms of 1000 m³.
· Fuelwood consumption in terms of 1000 m³ (Table A.1.FW).
· Charcoal consumption in terms of 1000 m³ (Table A.1.CH).

3.3.2.The IEA database

The data collecting approach

The International Energy Agency (IEA) publishes a yearbook on conventional energy statistics in countries belonging to the Organization for Economic Cooperation and Development (OECD). More recently, IEA has started to publish non-OECD country statistics, including biofuel consumption statistics.

Biofuel estimates were made on a country basis for 23 African countries. Thirty-one other countries were grouped together under "Other Africa"⁽¹⁰⁾, while St Helena was excluded from estimates. This classification assumed the use of a unified elasticity and base year for countries included under Other Africa.

The IEA is not officially authorized to collect data from non-OECD countries. Therefore, it adopts various approaches to complete its biofuel database. Data are primarily collected from various national sources such as publications, statistics, studies, or even surveys when available. When these primary sources are not available, the IEA collects data from regional organizations (e.g. the South African Development Community [SADC]) with which it has an information exchange and collaboration agreement. As a last resort, the IEA uses existing data at international level (UN statistics, FAO, African Development Bank, etc. Table 3.1 provides details on the different sources and parameters used by IEA in estimating biofuel use in each country.

In general, the original data sources used by IEA do not supply details on the origin of the data (surveys, derived data, estimated or simulated data, etc.). Therefore, it is impossible for IEA to provide quality assessment of the data and also for this study to evaluate the individual data on the IEA database.

In addition, since the collected data only generally present a few historical figures for each country (or even a single figure), a modeling approach had to be adopted to fill in the gaps. This approach reconstituted the time series for the period 1971-1996 for biomass TPES (Total Primary Supply of Combustible Renewable and Waste).

Biomass TPES estimates are based on a simple modeling approach that employs three main parameters to reconstitute the time series :

· Urbanization growth, which is a defined factor.

· A country-specific elasticity factor that links biomass TPES to urbanization growth. The estimates were made using a changing elasticity factor within three different reference periods: 1971-78, 1979-88 and 1989-95. These elasticity factors were defined by IEA using expert judgement of the potential interaction between biomass TPES and urbanization growth in each country concerned.

· Reference biomass TPES data for a given year, extracted from national sources or different regional and international sources. While considerable effort was made to collect national primary data, many biofuels estimates were also based on African Development Bank data or UN Statistics, which are themselves completely derived from FAO data. In addition, as mentioned above, data series were calculated individually for 23 countries, with 31 other countries grouped under "Other Africa" because of the lack of individual biomass TPES data in these countries.

As the model presents a historical series for aggregate biomass TPES supply during the period 1971-1996, it was difficult to reconstitute the specific contributions of different biomass types, such as woodfuels. However, detailed data presented for 1995 and 1996 allowed the specific contributions of fuelwood, charcoal, vegetal wastes and other biomass fuels to be distinguished and sectorial consumption to be divided up.

Structure of the database

One issue of the IEA publication including biomass data for 1995 is currently available, and the issue for 1996 is under publication.

These publications provide two types of data, all of which are presented in terms of TOE (tons of oil equivalent⁽¹¹⁾) :
· A time series on primary supply of combustible renewable and waste (Biomass TPES). These series cover data from 1971 to 1996, but do not provide detailed figures of the different wooduels considered or the sectorial consumption shares.
· Detailed data on biofuel use by type (fuelwood, charcoal, etc.) and by sector (transformation, energy and industry sector, residential, and non-specified) for the years 1995 and 1996.

Approach for using the data in the context of the study

Given the specific format of data presentation on the IEA database, it was not possible to present a complete historical series that fitted formatting requirements. Therefore, only data recorded for 1995 and 1996 were finally organized according to the study format:

1. Main tables :

· Fuelwood consumption for 1995 and 1996 in terms of m³, tons and PJ (Table A.2.1a / A.2.3a);
· Charcoal consumption for 1995 and 1996 in terms of m³ (primary wood used for carbonization), tons and PJ (Table A.2.1b / A.2.3b);
· Aggregate Woodfuel consumption for 1995 and 1996 in terms of m³, tons and PJ (Table A.2.1c / A.2.3c);
· Per capita woodfuel consumption for 1995 and 1996 in terms of m³ and tons (Table A.2.1);
· Primary conventional energy consumption in PJ (Table A.2.4);
· Direct and indirect woodfuel consumption in terms of m³ (Table A.2.5). However this table is not really relevant since IEA did not distinguish between direct and indirect woodfuel;
· Direct and indirect woodfuel consumption in energy terms : PJ (Table A.2.6). However this table is not relevant for the same reason as above;
· Fuelwood sectorial consumption in m³, tons and PJ (Table A.2.7a for 1995 and Table A.2.7b for 1996);
· Charcoal sectorial consumption in m³, tons and PJ (Table A.2.7c for 1995 and Table A.2.7d for 1996);
· Aggregate woodfuel consumption by sector in energy terms : PJ (Table A.2.8).

2. In addition, another original table was also included in the datafile : LPG consumption in tons (Table A.2.12).

Table 3.1 : Different sources and parameters used by IEA to estimate biofuel use in each country

	Data source used for Combustible Renewable and Wastes	Reference Year	Elasticities used for IEA estimates for the following range of years
			1971-78	1979-88	1989-95
Algeria	UN Statistics	1995	-0,50	-0,50	-0,50
Angola	ADB+SADC	1991	-0,30	-0,30	-0,30
Benin	ADB	1991	-0,10	-0,30	-0,30
Botswana	ADB+SADC	1991	-0,50	-0,50	-0,50
Burkina Faso	ADB	1991	-0,50	-0,50	-0,50
Burundi	ADB+ESMAP	1991	-0,50	-0,50	-0,50
Cameroon	ADB	1991	-0,10	-0,10	-0,10
Cape Verde	ADB	1991	-0,50	-0,50	-0,50
Central Africa Republic	ADB	1991	-0,50	-0,50	-0,50
Chad	ADB	1991	-0,50	-0,50	-0,50
Comoros	ADB	1991	-0,50	-0,50	-0,50
Congo Dem. Rep.	ADB	1991
Congo	ADB	1991	-0,30	-0,30	-0,30
Cote d'Ivoire	ADB	1994	-0,20	-0,20
Djibouti	ADB	1991	-0,50	-0,50	-0,50
Egypt	UN Statistics	1995	-0,50	-0,50	-0,50
Equatorial Guinea	ADB	1991	-0,50	-0,50	-0,50
Eritrea	National Sources	1994	-0,50	-0,50	-0,50
Ethiopia	AFREPREN	1992	-0,10	-0,10	-0,10
Gabon	ADB + National Sources	1991	-0,30	-0,30	-0,30
Gambia	ADB		-0,50	-0,50	-0,50
Ghana	National Sources	1994	2,50	-	-
Guinea	ADB	1991	-0,50	-0,50	-0,50
Guinea-Bissau	ADB	1991	-0,50	-0,50	-0,50
Kenya	ADB	1991	-0,30	-0,30	-0,30
Lesotho	ADB+SADC	1991	-0,50	-0,50	-0,50
Liberia	ADB	1991	-0,50	-0,50	-0,50
Libya	UN Statistics	1995	-0,50	-0,50	-0,50
Madagascar	ADB	1991	-0,50	-0,50	-0,50
Malawi	ADB+SADC	1991	-0,50	-0,50	-0,50
Mali	ESMAP	1987	-0,50	-0,50	-0,50
Mauritania	ADB	1991	-0,50	-0,50	-0,50
Mauritius	National Sources	1994	-0,50	-0,50	-0,50
Morocco	UN Statistics	1995	-0,50	-0,50	-0,50
Mozambique	ADB+SADC	1991	-0,10	-0,20	-0,20
Namibia	ADB + SDC	1991	-0,50	-0,50	-0,50
Niger	ADB	1991	-0,50	-0,50	-0,50
Nigeria	ADB	1991	-0,10	-0,10	-0,10
Reunion	ADB	1991	-0,30	-0,30	-0,30
Rwanda	AFREPREN	1991	-0,50	-0,50	-0,50
Sao Tome and Principe	ADB	1991	-0,50	-0,50	-0,50
Senegal	ADB + Enda + National Sources	1994	-0,30	-0,30	-0,30
Seychelles	ADB	1991	-0,50	-0,50	-0,50
Sierra Leone	ADB	1991	-0,50	-0,50	-0,50
Somalia	ADB	1991	-0,50	-0,50	-0,50
South Africa	National	1995	-0,30	-0,30	-0,50
Sudan	AFREPREN	1990	-0,30	-0,30	-0,30
Swaziland	ADB+SADC	1991	-0,50	-0,50	-0,50
Tanzania	SADC	1990	-0,30	-0,30	-0,30
Togo	ADB	1991	-0,50	-0,50	-0,50
Tunisia	National Sources	1995	-0,10	-0,10	-0,10
Uganda	National Sources	1991	-0,50	-0,50	-0,50
Zambia	ADB+SADC	1991	-0,30	-0,30	-0,30
Zimbabwe	ADB+SADC	1991	-0,30	-0,30	-0,30
Other Africa		1995	-0,10	-0,10	-0,10

3.3.3. ESMAP statistics

The Energy Sector Management Assistance Programme (ESMAP), jointly executed by the World Bank and UNDP, did not aggregate any of the energy data (especially for woodfuels) that were collected by the numerous projects initiated in various countries.

However, since almost all the published ESMAP documents for African countries were available in the FAO Library, they were consulted, compiled and assembled for the current study.

In the 1980s, ESMAP published two types of documents: energy policy issues, and final reports of specific ESMAP projects (particularly those expressly addressing household energy strategy).

The data collecting approach

For the first type of document, expert missions were undertaken in order to present the state of the art in the energy sector and to identify relevant areas of intervention. While no particular effort was made to consolidate woodfuel data, which were simply included in the energy balances (with no specifications on data origin and quality), these missions made a substantial effort to compile the best national data.

As regards project reports, many ESMAP projects addressed household energy issues, and substantial information was generated from field surveys; scope varied according to the specific issues identified in each of the concerned countries (e.g. urban areas, specific cities, etc.).

In both cases, and particularly for projects, ESMAP documents presented the best estimates available at that time. Therefore, the data could be used as a good basis for filling in gaps and reconstituting the time series.

ESMAP documents for 47 countries were consulted for this current study. The woodfuel data compiled and assessed are presented in Appendix 3.3.

Structure of ESMAP statistics

In general, ESMAP documents reported woodfuel data as a part of the energy balances in terms of ton of oil equivalent (TOE). However data were also sometimes available in terms of tons or cubic meters.

On the other hand, the data were presented for different periods, depending on data availability at national levels or on the year of the survey undertaken by ESMAP projects. In general, time series were not available except when the ESMAP teams were able to undertake simulation activity. However such derived figures cannot considered as concrete data for the purpose of this study.

Approach for using the data in the context of the study

Generally the ESMAP studies employed the existing definition of woodfuel uses, which is the same as the FAO definition. However, ESMAP generally applied specific conversion factors for fuelwood and charcoal, some of which differed from one country to another.

Where possible, data collected from ESMAP documents were included in the same eight tables used for the other data sources. However the data were only included when the conversion factors of woodfuels to original units (e.g. metric tons or m³) were clearly defined.

Finally, the following tables were built up :

1. Figures of aggregate woodfuel consumption (from 1970 to 1990, where data were available):

· In terms of volume (Table A.3.1 : woodfuel consumption 1000 m³, and Table A.3.1 Per capita woodfuel consumption in terms of m³/year);

· In energy terms (Table A.3.3 : Petajoules);

2. Figures presenting :

· Direct and indirect woodfuel consumption (Table A.3.5a and Table A.3.5b) in terms of m³;

· Direct and indirect woodfuel consumption (Table A.3.6) in energy terms (PJ).

3. Figures of sectorial woodfuel consumption (where data were available) :

· In terms of original units (Table A.3.7a : Fuelwood consumption 1000 tons, Table A.3.7b : Charcoal consumption 1000 tons, Table A.3.7c : Aggregated woodfuel consumption 1000 tons);

· In primary energy terms (Table A.3.8 : Petajoules);

4. In addition, original tables were also presented using a similar format but in the original units :

· Aggregate final woodfuel consumption in energy terms : PJ (Table A.3.14)

· Aggregate woodfuel consumption in terms of tons (Table A.3.17)

3.3.4. The ENDA statistics

The data collecting approach

Jointly with the Institut de l'Economie et Politique de l'Energie (IEPE, Grenoble, France), ENDA (Senegal) issued a synoptic study on the energy situation in Francophone African countries in 1995. This study aimed at compiling existing data on different energy commodities including biofuels. Except for Senegal, where some annual data on woodfuels (fuelwood and charcoal) were presented, the document does not present either data series or detailed figures on sectorial woodfuel consumption.

However, the data were collected at national level and thus could be consistently used in defining reference data for some countries where information was uncertain or unavailable from other sources.

Structure of the ENDA statistics

In general, the ENDA/IEPE document reports woodfuel in a simple format, providing figures on total fuelwood and charcoal consumption for each country studied in terms of metric tons. No details are provided on sectorial share, area shares (urban versus rural), and no comments are offered on data sources (surveys, studies, official documents, international data, etc.), data quality (primary, derived or estimated data), and the conversion factors to be applied.

Approach for using the data in the context of the study

The tables for fuelwood and charcoal consumption are presented using original units (1000 tons). Aggregate figures (fuelwood + charcoal) are presented using a similar energy conversion factor as FAO⁽¹²⁾. However, in order to present comparable figures, these data were converted into 1000 m³ and PJ according to the same format as the other statistics.
Finally the following uniform tables were presented:
3. Figures of global woodfuel consumption (from 1985 to 1994, where data were available) :

· In terms of volume (Table A.4.1 : woodfuel consumption 1000 m³, and Table A.4.2 Per capita woodfuel consumption : m³/year)

· In energy terms (Table A.4.3 : Petajoules).

4. In addition, original tables were also presented using a similar format but with the original units :

· Fuelwood consumption in terms of 1000 tons (Table A.4.9)

· Charcoal consumption in terms of 1000 tons (Table A.4.10)

· Aggregate woodfuel consumption in terms of 1000 tons (Table A.4.11)

· LPG consumption in terms of 1000 tons (Table A.4.12)

· Final conventional energy consumption in terms of 1000 TOE (Table A.4.13)

· Final woodfuel consumption of (Table A.4.14PJ)

· Final consumption of woodfuels as a function of the final energy consumption in terms of percentage (Table A.4.14 (%))

3.3.5. Other national woodfuel statistics

More than 120 different national, regional and international documents and studies addressing energy issues, particularly woodfuels, in African countries were consulted and compiled for this study. These data are presented in Appendix 3.5 for different time periods, and elements for cross-checking all data are provided.

The data collecting approach

Unlike the other formal databases or documents, there was no specific approach for addressing woodfuel data. The limited scope and duration of the study have provided only a few possibilities to collect data from different sources, including :

· various documents from the FAO Library;

· various documents from the IEA Library;

· various documents from the consultant's library;

· various documents from web sites (only Senegal has a relevant web site).

However, although no national institutions were contacted during the current study, the study undertook an unprecedented collecting and compiling effort, and almost all available documents in the specified libraries were consulted and adapted for inclusion in the different tables of Appendix 3.5, even those which were not systematically consistent. While certain relevant national documents may have been missed, it is unlikely that the analysis presented would have been modified significantly, at least at continental and sub-regional levels.

The ongoing regional workshops to be organized by FAO in 1999 in Africa certainly represent a unique opportunity to confirm this assumption, to consolidate the state-of-art of available data definitively, particularly at national levels, and to update knowledge on woodfuel topics in Africa.

Structure of the other national woodfuel statistics

In general, the collected woodfuel data were originally presented in various ways (aggregate woodfuels, specific fuelwood and charcoal consumption); units (TOE, tons, m³, etc.) and; level of details (aggregate consumption, sectorial demand, consumption by area, etc.).

Except in a few cases, the data sources did not document the data origin (primary data from surveys, derived or estimated data, etc.), and did not address data quality issues. In addition, data sources tended to focus on woodfuel demand and only a few sources addressed woodfuel supply.

Approach for using the data in the context of the study

Data extracted from the various documents consulted were originally presented in different formats, units, time horizons and scope. Although sometimes inconsistent, all data are included in Annex 3.5, allowing the reader to check and be aware of the wide range of estimates available.

The most frequent formats examined presented national consumption in terms of tons, m³ or other energy specific unit (TOE, PJ, etc.). Less frequently, certain documents presented sectorial woodfuel consumption, and in certain cases distinguished between fuelwood, charcoal and other woody biomass. The main difficulty was to interpret the figures collected consistently. However, this was not always easy; in many cases, it was not clear whether the reported figures covered all types of woodfuels, all consuming sectors, all areas (rural and urban), etc. Moreover, relevant details such as conversion factors, specific carbonization efficiencies, etc., were not systematically reported, and almost none of the documents indicated whether data were primary (field surveys) or secondary (available national estimates, estimates made by the authors, simulations based on models, etc.).

However, in order to conserve the significance of the original data, only the data that could be presented suitably using the defined unified format were reported. Therefore, many approximate data were excluded from the analysis, such as figures for woodfuel consumption derived from per capita woodfuel consumption without any detailed explanation (e.g. sectors included, woodfuels included, areas included, corresponding year of the data, etc.).

While considerable effort was made to present the original data faithfully, certain initiatives were taken to fill in the gaps. The general approach was to adopt the FAO definitions when some parameters were missing. For example, different assumed FAO parameters (e.g. calorific values of different woodfuels, moisture contents, etc.) were used unless specific and in

Ultimately, data were presented in the unified format used for the other data sources (FAO, IEA, etc.). The original data were included in the following two files:

1. Figures of sectorial fuelwood consumption (from 1973 to 1997, where data were available) in terms of 1000 tons (Table A.5.7a : Fuelwood consumption);

2. Figures of sectorial charcoal consumption (from 1973 to 1997, where data were available) in terms of 1000 tons (Table A.5.7b : Fuelwood consumption);

All other tables that were directly derived from the two tables include:

· Sectorial woodfuel consumption (from 1973 to 1997) in terms of 1000 m³ (Table A.5.7c : Woodfuel consumption);

· Total woodfuel consumption in terms of 1000 m³ (Table A.5.1);

· Per capita woodfuel consumption in terms of m³/year (Table A.5.1 PerCapita);

· Total national woodfuel consumption in energy terms : Petajoules (Table A.5.3);

· National woodfuel consumption dividing direct and indirect woodfuel use in terms of 1000 m³ (Table A.5.4). It should be recalled that Cote d'Ivoire was the only country where indirect woodfuel consumption was reported in national literature. This obviously does not mean that it is the only country that uses indirect woodfuel sources.

· National woodfuel consumption dividing direct and indirect woodfuel use in energy terms : Petajoules (Table A.5.6);

· Sectorial woodfuel consumption in energy terms : Petajoules (Table A.5.8);

· Aggregate woodfuel final consumption in energy terms : Petajoules (Table A.5.14PJ);

· Aggregate woodfuel consumption in terms of 1000 Tons (Table A.5.17).

3.3.6. Other international woodfuel statistics

African Development Bank statistics

In 1994, the African Development Bank (ADB) launched an African Energy Program, which carried out different expert missions in many African countries. The ADB programme addressed woodfuel issues and presented available data. These data were derived from the FAO database with no particular effort made to assess them or to compile any other data. Therefore no specific analysis or presentation of ADB data are made in the current report.

United Nations statistics

The annual United Nations publication "Energy Statistics Yearbook" presents statistical time series on the production, trade and consumption of energy, including biofuels. However, woodfuel data are completely derived from FAO statistics, the only difference being in the use of different conversion factors. Therefore the UN data were not analysed or presented in great detail in this report.

The World Resources Institute (WRI) statistics

The annual publication of the World Resources Institute "World Resources" presents statistical time series on energy production and consumption, including biofuels. However, again the data on woodfuels are also completely derived from FAO statistics. Therefore, no specific analysis or presentation of the WRI data are made.

3.4.Comparison of the aggregated results of the different databases and statistics

3.4.1. Different possibilities for comparison

It is not easy to compare data derived from a variety of sources. In fact, only a few comparable elements can be reported consistently. Difficulties arise for the following reasons:

· The FAO database is the only data source that provides almost comprehensive figures (all countries are included except 6 non woodfuel-determining countries : Cape Verde, Djibouti, Sao Tome and Principe, Saint Helena, Comoros, and Seychelles). In addition the time series are continuous for the whole scope of the study: 1980-1996. However, the FAO database only presents figures estimated by using assumed per capita consumption rather than actual figures. Furthermore FAO statistics provide no sectorial consumption figures.

· The IEA database presents individual data for only 23 countries. In order to complete the figures, 31 countries are grouped under "Other Africa" and their corresponding data are estimated approximately, while data on Saint Helena are not considered. In addition, sectorial woodfuel consumption is only available for 1995 and 1996, and data for the remaining years are only considered globally (data are estimated in terms of TOE of Primary Supply of Combustible Renewable and Waste, making it difficult to retrieve the corresponding figures for woodfuels considered under this item for each country).

· The ESMAP documents considered data for 39 countries out of 55. In fact, no data were reported for Chad, Djibouti, Eritrea, Guinea, Cameroon, Central African Republic, Equatorial Guinea, Botswana, Saint Helena, Zimbabwe, Mauritius, Reunion, Algeria, Egypt, Libya, and South Africa. In addition, the ESMAP documents presented data of inconsistent quality and reported data for more than one reference year in only ten countries.

· The ENDA/IEPE documents reported data for only 28 countries, only 3 of which are not Francophone countries. In addition, except for Senegal, data were presented for only one year and no detailed sectorial consumption data were provided.

· Other national data collected were included in the tables as found. Therefore, it was not possible to aggregate the figures at sub-regional or African level, or to build up time series before undertaking an in-depth review of the data. This review was included within the major component of this study, which was to finalize the best estimates. In addition, data were missing for 15 countries : Djibouti, Eritrea, Benin, Burundi, Cameroon, Central African Republic, Congo Democratic Republic, Equatorial Guinea, Gabon, Rwanda, Sao Tome and Principe, Saint Helena, Comoros, Madagascar, Reunion, and Libya.

Nevertheless, it might be interesting to show the gaps and the quality levels of the available data. This could help in identifying areas for data improvement and ways to achieve this.

3.4.2. Aggregate woodfuel consumption

Comparative analysis of aggregate woodfuel consumption figures from different sources was not relevant, particularly since these sources did not cover the same number of countries and provided incomplete time series. Therefore, the data presented in the following tables should only be viewed as illustrating the gaps characterizing woodfuel data in Africa.

Table 3.2 : Aggregate Woodfuel Consumption in Africa According to the Different Data Sources and Different Reference Years (1000 m³ )

Regions	FAO 1990[1]	IEA 1996[2]	ESMAP 1981[3]	Enda/IEPE 1989[4]	Others 1990[5]
West Sahelian Africa	27 328	5 765	7 476	14 469	17 145
East Sahelian Africa	95 264	126 242	57 299	-	87 956
West Moist Africa	130 769	173 125	2 542	23 992	128 592
Central Africa	77 469	68 911	0	14 076	25 457
Tropical Southern Africa	78 656	115 829	54 142	20 527	108 452
Insular East Africa	7 724	-	0	9 607	0
North Africa	9 392	10 625	19 786	8 571	11 600
Non Tropical Southern Africa	10 807	63 177	39	-	3 088
Other Africa	-	119 947
Total Africa	437 409	683 622	141 285	91 241	382 290

[1] Data are available for 47 countries.
[2] Data are available for 23 individual countries, 31 other countries being grouped under OTHER AFRICA. Therefore, regional figures were not relevant. In addition, woodfuel data by item are provided only for 1995 and 1996.
[3] 1981 is the best year in terms of data availability (the figures include 10 countries' data)
[4] 1989 is the best year in terms of data availability (the figures include 17 countries' data)
[5] In 1990, the figures included 22 countries' data

3.4.3. The sectorial share of woodfuel consumption

Again it is not really relevant to compare data derived from different sources. However, by way of illustration, it might be interesting to underline the important share of households in the woodfuel balance for those sources where sectorial figures were provided: IEA, ESMAP and Others :

Table 3.3 : Sectorial Woodfuel Share in Africa According to the Different Data Sources and Different Reference Years (%)

Final Energy Users	IEA 1996	ESMAP 1981[1]	Others 1990
Households	87 %	95.6 %	95.9 %
Industries	11 %	2.1 %	1.7 %
Transformation
Other sectors	2 %	2.3 %	2.4 %
Total	100%	100%	100%

3.4.4. Per capita woodfuel consumption

Comparison of per capita consumption among the different data sources might prove more consistent despite the gaps. In fact, the aggregate sub-regional averages are much less affected by these gaps given the general similarities of global woodfuel behaviour within a limited geographical area.

The numerous gaps, particularly in IEA, ESMAP and ENDA data, prevented consistent sub-regional comparisons. However, these gaps may not have systematically affected comparability of the aggregate African average from these sources. In fact, the following table shows relatively comparable aggregated figures, particularly between FAO and ENDA sources in one side, and between IEA, ESMAP and other national sources, on the other.

Table 3.4 : Per Capita Woodfuel Consumption in Africa According
to the Different Data Sources (m³/year)

Regions	FAO 1990	IEA 1996	ESMAP 1981[2]	Enda/IEPE 1989[3]	Others 1990[5]
West Sahelian Africa	0.639	0.676	0.616	0.567	0.710
East Sahelian Africa	0.889	1.114	2.978	-	1.181
West Moist Africa	0.912	1.136	0.945	0.968	1.014
Central Africa	0.918	1.074	-	0.851	1.348
Tropical Southern Africa	1.009	1.457	1.444	0.898	1.392
Insular East Africa	0.540	NA	-	0.781	0.000
North Africa	0.080	0.079	0.998	0.364	0.482
Non Tropical Southern Africa	0.273	1.490*	0.028	-	1.222
Other Africa		0.844
Total Africa	0.697	0.929	1.523	0.727	1.096

(*) Data are available only for South Africa

6 "x"refers to the origin of the data presented in every table (1 =FAOSTAT, 2=IEA, 3= ESMAP, 4=ENDA/IEPE, 5=Oth_Nat) and "y"refers to woodfuel commodities.

7 m³= CUM

8 1 Petajoule (PJ) = 1015 Joule).

9 According to FAO assumptions, 1 ton of charcoal is derived from 6 m3 of wood.

10 Refer to Appendix 3.2 (IEA tables)

11 The conversion factors used by IEA were 13.7 Mj/kg (or 0.328 toe/ton) for fuelwood and 30.7 Mj/kg (or 0.733 toe/ton) for charcoal.

12 13.8 Mj/kg of fuelwood and 30.8 Mj/kg of charcoal. While the same FAO wood density was assumed (1 m3 of wood = 0.725 ton), a more realistic wood/charcoal ratio was applied (6 tons wood for 1 ton charcoal).