Scholarly article on topic 'Additional cooking fuel supply and reduced global warming potential from recycling charcoal dust into charcoal briquette in Kenya'

Additional cooking fuel supply and reduced global warming potential from recycling charcoal dust into charcoal briquette in Kenya Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Journal of Cleaner Production
Keywords
{"Climate impact" / "Cooking fuel" / Kenya / "Life cycle assessment"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — M. Njenga, N. Karanja, H. Karlsson, R. Jamnadass, M. Iiyama, et al.

Abstract Rising demand for energy is one of the major challenges facing the world today and charcoal is a principal fuel in Kenya. Faced with energy poverty many poor households turn to briquette making. This study assessed the additional cooking fuel obtained from recycling charcoal dust into charcoal briquettes. It applied Life Cycle Assessment (LCA) to assess the global warming potential (GWP) from use of charcoal and production of briquettes from charcoal dust and cooking a traditional meal for a standard household of five people. Native vegetation of Acacia drepanolobium and a low efficiency kiln were considered the common practice, while an Acacia mearnsii plantation and a high efficiency kiln was used as an alternative scenario. Charcoal and kerosene were considered as reference fuels. Recovering charcoal dust for charcoal briquettes supplied an additional 16% cooking fuel. Wood carbonization and cooking caused the highest GWP, so there is a need for technologies to improve the efficiency at these two stages of charcoal briquettes and charcoal supply chain. Supplying energy and cooking a traditional meal in a combined system using charcoal and recovering charcoal dust for charcoal briquettes and charcoal alone accounted for 5.3–4.12 and 6.4–4.94 kg CO2 eq. per meal, respectively, assuming trees were not replanted. These amounts declined three times when the carbon dioxide from the carbonization and cooking stages was assumed to be taken up by growing biomass. This requires replanting of trees cut down for charcoal if the neutral impact of biomass energy on GWP is to be maintained.

Academic research paper on topic "Additional cooking fuel supply and reduced global warming potential from recycling charcoal dust into charcoal briquette in Kenya"

ELSEVIER

Additional cooking fuel supply and reduced global warming potential from recycling charcoal dust into charcoal briquette in Kenya

M. Njenga a'b' *, N. Karanja a, H. Karlsson c, R. Jamnadass b, M. Iiyama b, J. Kithinji a, C. Sundberg c

a University of Nairobi, Box 30197-00100, Nairobi, Kenya

b World Agroforestry Centre, Box 30677-00100, Nairobi, Kenya

c Swedish University of Agricultural Sciences, Box 7032, 750 07 Uppsala, Sweden

ARTICLE INFO ABSTRACT

Rising demand for energy is one of the major challenges facing the world today and charcoal is a principal fuel in Kenya. Faced with energy poverty many poor households turn to briquette making. This study assessed the additional cooking fuel obtained from recycling charcoal dust into charcoal briquettes. It applied Life Cycle Assessment (LCA) to assess the global warming potential (GWP) from use of charcoal and production of briquettes from charcoal dust and cooking a traditional meal for a standard household of five people. Native vegetation of Acacia drepanolobium and a low efficiency kiln were considered the common practice, while an Acacia mearnsii plantation and a high efficiency kiln was used as an alternative scenario. Charcoal and kerosene were considered as reference fuels. Recovering charcoal dust for charcoal briquettes supplied an additional 16% cooking fuel. Wood carbonization and cooking caused the highest GWP, so there is a need for technologies to improve the efficiency at these two stages of charcoal briquettes and charcoal supply chain. Supplying energy and cooking a traditional meal in a combined system using charcoal and recovering charcoal dust for charcoal briquettes and charcoal alone accounted for 5.3—4.12 and 6.4—4.94 kg CO2 eq. per meal, respectively, assuming trees were not replanted. These amounts declined three times when the carbon dioxide from the carbonization and cooking stages was assumed to be taken up by growing biomass. This requires replanting of trees cut down for charcoal if the neutral impact of biomass energy on GWP is to be maintained.

© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

CrossMark

Article history: Received 18 April 2013 Received in revised form 29 May 2014 Accepted 2 June 2014 Available online 12 June 2014

Keywords: Climate impact Cooking fuel Kenya

Life cycle assessment

1. Introduction

Rising energy demand is one of the major challenges facing the world today. About 2.4 billion people use solid biomass fuels as a source of energy for cooking and heating (Kaygusuz, 2011). In sub-Saharan Africa (SSA), wood-based fuels account for over 80% of primary energy supply and more than 90% of the population rely on firewood and charcoal (IEA, 2006). Charcoal is the principal fuel in Kenya (Karekezi, 2002), providing energy for 82% of urban and 34% of rural households (MoE, 2002). Kituyi (2004) argues that for the short and medium term, any sustainable development solutions in

* Corresponding author. World Agroforestry Centre, Box 30677-00100, Nairobi, Kenya. Tel.: +254 (0) 722 331 006.

E-mail addresses: m.njenga@cgiar.org (M. Njenga), n.karanja@cgiar.org (N. Karanja), hanna.e.karlsson@slu.se (H. Karlsson), r.jamnadass@cgiar.org (R. Jamnadass), m.iiyama@cgiar.org (M. Iiyama), jkithinji@uonbi.ac.ke (J. Kithinji), Cecilia.Sundberg@slu.se (C. Sundberg).

the household energy sub-sector in Africa must focus on biomass energy technology development and dissemination. Kerosene is used by approximately 93% of households for both cooking and lighting by the urban population while in the rural areas, the principal use is for lighting (MoE, 2002). Kerosene is mainly produced at a refinery in Mombasa.

Most of Kenya's charcoal producers (86%) source wood from private farms owned either individually or communally, while the rest is from government or county council land. Most of these charcoal sources are woody savannah, which covers over two-thirds of the country's area (MoE, 2002; Mutimba and Barasa, 2005). The wood found in the dry savannah is usually hard, dense and with a low moisture content, yielding good quality charcoal. Charcoal is produced by heating fuelwood in some type of kiln with limited access to air, a process called carbonization which creates a fuel with higher energy content than air-dried fuelwood (Pennise et al., 2001).

A key environmental impact to consider when assessing the sustainability of energy sources is the climate impact over the life

http://dx.doi.org/10.1016/j.jclepro.2014.06.002

0959-6526/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

cycle of energy supply and use. The climate impact from emission of greenhouse gases (GHG) is commonly quantified as global warming potential (GWP), expressed in carbon dioxide equivalents (CO2-eq). The resulting GWP of wood fuels depends strongly on whether trees are assumed to regrow, i.e. whether or not the CO2 emitted when burning causes a net increase in atmospheric CO2.

Traditional kilns are favoured across SSA because they require very little capital investment, are flexible in size and shape, and are mobile. Hence they are well-matched to the dispersed nature of the charcoal trade (Mugo and Poulstrup, 2003). However, these traditional kiln have low efficiency in converting wood to charcoal (Mutimba and Barasa, 2005; Okello et al., 2001). There are improved kilns with higher efficiency and better charcoal quality, but improved production techniques require more labour and have higher costs (Oduor et al., 2006). There are no incentives or credit facilities to enable communities to construct these improved kilns, while poor implementation of charcoal policies has inhibited development of the charcoal trade.

Because of inherent inefficiencies in the carbonization process, there is a substantial loss of carbon and energy from the starting fuelwood, primarily as carbon dioxide (CO2) but also products of incomplete combustion (PIC) such as carbon monoxide (CO), methane (CH4), particulate matter (PM) and nitrous oxide (N2O) (Pennise et al., 2001). CO indirectly affects global warming through atmospheric photochemical reactions that in turn affect GHG levels. CH4 and CO have higher GWP per kilogram (kg) of carbon than CO2 (IPCC, 2007). Emissions of many GHG from cooking stoves are the result of the significant proportion of fuel carbon that is diverted to PIC as a result of poor combustion efficiency (Edwards et al., 2003).

Faced with the challenges of poverty, unemployment and access to cooking fuel, many poor households turn to briquette making. In Kenya, as in many other developing countries, briquette production is focused on providing good quality cooking fuel. In the slum Kibera, Nairobi, 70% of households living within 250 m of a briquette production site used charcoal briquettes, thereby saving over 50% on the cost of cooking fuel (Njenga et al., 2013a). Briquettes are mainly used as a substitute for charcoal, as both use similar cooking stoves (Aya et al., forthcoming). Half the briquette enterprises in Kenya use charcoal dust as the main raw material, sourced from charcoal sellers in urban areas and forming 10—15% of the charcoal supply chain. Around 50% of briquetting activities are by community-based groups and Nairobi city hosts half of these (Terra Nuova et al., 2007). Provision of cooking energy from a range of sources to meet people's needs will require adequate, reliable and affordable supplies that result in minimal impacts on the environment (Olz et al., 2007). It is therefore important to establish the environmental impacts of charcoal briquettes, in order to inform decision making in developing sustainable cooking biomass fuel to meet one of Africa's greatest sustainability challenges: energy insecurity. In a study on the charcoal supply chain in Kenya, Kituyi (2004) found that applying Life Cycle Management (LCM) can deliver social, economic and environmental benefits to developing country communities and should therefore be promoted. A previous study on emissions from charcoal making in Kenya recommended a full analysis of the charcoal life cycle, including evaluation of its final end use in combustion in cooking stoves (Pennise et al., 2001). In another LCA on charcoal, biogas and liquid petroleum gas (LPG) in Ghana by Afrane and Ntiamoah (2011), global warming and human toxicity were the most significant overall environmental impacts associated with use of these fuels, with charcoal and LPG, respectively, making the largest contribution to these impact categories.

The aim of the present study was to quantify the potential benefits from recycling charcoal dust into charcoal briquettes

regarding reduction of GHG emissions and increased cooking fuel supply. The GHG emissions were quantified using LCA methodology. The study was carried out to address the above-mentioned knowledge gaps, including determination of emissions factors from cooking stoves using charcoal briquettes, charcoal and kerosene, and assessment of the GWP from cooking a traditional meal for a standard household of five people. Furthermore, the additional amount of cooking fuel produced from recycling charcoal dust into charcoal briquettes was determined. The study also assessed the wood and land requirements for wood production for charcoal and charcoal briquettes in different production systems.

2. Materials and methods

The study on GWP was carried out in accordance with ISO 14044 (2006), which specifies requirements and guidelines for conducting LCA.

2.1. Goal and scope of the study

2.1.1. Goal, functional unit and comparisons made

The main goal of the study was to quantify the GWP in different stages of the production and use of charcoal briquettes as a cooking fuel in Kenya. The LCA aimed at identifying stages that will require technological and policy interventions in the life cycle of charcoal briquettes to promote its development as a sustainable cooking fuel. Another goal was to establish emission factors (EF) for CO, CO2 and fine PM (PM25) from households cooking with charcoal briquettes, charcoal and kerosene. The functional unit was fuel used in cooking a traditional meal - a mixture of 500 grams (g) of green maize (Zea mays) and 500 g of dry common bean (Phaseolus vul-garis) commonly known as githeri — for a standard Kenyan household of five people (Kenya Government, 2010). Climate impact as GWP100 was calculated in kg CO2 equivalent (CO2 eq) per meal, using SimaPro software (SimaPro 7.3.3, 2011). Input data for the life cycle inventory analysis were based on our own measurements, literature values and the Ecoinvent database (Ecoinvent Centre, 2010). The distinct processes in the LCA included common practices of sourcing wood from Acacia drepanolobium native woodland savannah, wood carbonization using a low efficiency earth mound kiln, transportation of charcoal to urban areas and use of charcoal briquette in cooking (Fig. 1 ). In addition, comparisons were made with improved practices such as wood production in an Acacia mearnsii plantation and wood carbonization using a high efficiency mound kiln. Furthermore, cooking the standard meal with two reference fuels, charcoal and kerosene, was assessed, as the former is the most common in urban areas in SSA, while the latter is the fossil fuel that is most widely used in Kenyan households.

2.1.2. System description

2.1.2.1. Systems. Two charcoal production systems were studied:

i. Common or traditional system: A. drepanolobium native woodland savannah and a low efficiency carbonization process.

ii. Alternative or improved system: A. mearnsii plantation and a

high efficiency carbonization process.

For both charcoal production systems, two fuel use systems were compared:

a. Use of charcoal and production of briquettes from charcoal dust

b. Use of charcoal without use of charcoal dust

Common practice for charcoal briquettes

Improved scenario for charcoal briquettes

Acacia drepanolobium native woodland

Wood carbonization with low efficiency kiln

Acacia mearnsii plantation

Wood carbonization with high efficiency kiln

Charcoal transportation by lorry

Sourcing water from river

Briquette production using charcoal dust

Sourcing of soil from river banks and road reserves

Household use of charcoal briquettes

Fig. 1. Flowchart of charcoal briquette production and use in Kenya.

2.1.2.2. Wood production. The two wood production systems, an A. drepanolobium woodland savannah system and an A. mearnsii plantation system, were compared under two different assumptions: 1) That wood production was not renewable, i.e. that biomass would not regrow; and 2) that wood production was renewable and carbon-neutral, based on natural vegetation regrowth. The production of wood from native A. drepanolobium vegetation on a private farm was considered as the common practice (Fig. 1). A. drepanolobium is an ideal candidate for sustained charcoal production because: (a) it occurs in almost mono-specific stands in high densities over vast areas; (b) it coppices readily when harvested or top-killed by fire; (c) its hard wood makes good quality charcoal; and (d) income from its charcoal is an attractive source of supplementary revenue. Under the A. drepanolobium woody savannah system, charcoal producers harvest mature stems and leave the young ones to grow from the same plant. This native vegetation system was compared with production of wood from an A. mearnsii plantation as the improved case. Under the A. mearnsii plantation system, mature stems are harvested and the young ones left to grow for next harvesting, and the land is always covered by vegetation. In the native vegetation system a 14-year rotational cycle and in the plantation system a 9-year rotational cycle were considered, with yield of biomass usable for charcoal production of 18.3 tonnes per hectare (t/ha) and 65 t/ha, respectively (Okello et al., 2001; Cheboiwo and Mugo, 2011). No inputs of fossil fuels or fertilizer were assumed in the plantation system. Wood was

harvested manually and transported by manpower to the carbonization site using wheelbarrows in both systems.

In the case of charcoal, both the current practices and improved scenarios were considered up to transportation of the charcoal, as shown in Fig. 1. After that, the charcoal was assumed to be used as cooking fuel and the charcoal dust burned in the open at the charcoal trading places.

21.2.3. Wood carbonization and transportation of charcoal. To determine the amount of wood required to be carbonized into charcoal to cook a traditional meal, a traditional earth mound kiln with efficiency in yield of 14% dry mass was assumed, as it is the most common practice used by charcoal producers in Kenya (Okello et al., 2001; Mutimba and Barasa, 2005) (Fig. 2a).

For comparison, a high efficiency mound kiln of the Masonry brand with an efficiency of 33% being used by Kakuzi Ltd and the former EATECH Eldoret was considered (Kakuzi, 2003) (Fig. 2b).

In the calculations of GHG emissions from wood carbonization, emission factors for an improved kiln and the traditional earth mound kilns in Kenya reported by Pennise et al. (2001) were applied. The emission factors for the wood carbonization processes are presented in Table .

After charcoal is produced, it is packed into recycled sacks originally used to pack produce such as sugar, maize and maize flour. Each sack contains about 40 kg of charcoal Transportation of charcoal from the production site to the road is done using donkey

Fig. 2. (a) Low efficiency earth mound kiln (Photo by Nelly Oduor). (b) High efficiency Masonry mound kiln (Photo by Mary Njenga).

Table 1

Emission factors (g pollutant/kg charcoal produced) Source: Pennise et al. (2001).

Description of efficiency CO2 CO CH4 NOx N2O

aLow efficiency earth 2510 270 40.7 0.109 0.21

mound kiln

bHigh efficiency Masonry 1103 169 47.0 0.033 0.076

mound kiln

a Average of EM1 and EM2 = traditional earth kilns with efficiency 22% replicates 1 and 2.

b Average of EM4 and EM5 = an improved kiln with 33% efficiency.

carts, which were not included in the calculation. Transportation distance to Nairobi city, one of the major consumers of charcoal, was taken as 200 km, which is the radius for sources of charcoal such as the counties of Narok, Machakos, Laikipia and Kajiodo. Data on a 'lorry 3.5—7.5t, EURO3/RER S' was selected from the Ecoinvent database as charcoal is transported by individual traders who use smaller trucks.

2.1.2.4. Waste management of charcoal dust and its recycling into charcoal briquettes. Charcoal dust, which comprises 10—15% (average of 13% assumed) in the Kenyan charcoal supply chain, is mainly found at the charcoal selling places and is burned or disposed of in open drains (Mugo et al., 2007). In the charcoal briquette case, the charcoal dust is used in charcoal briquette production (Fig. 1). In the charcoal reference case, burning was considered the common practice of managing this waste and emissions assumed to be the same as those emitted during burning of charcoal in a kitchen as described below. The charcoal briquette studied represents a type made by community groups in Nairobi. The local people mix charcoal dust with soil as a binder at relative proportions of 4:1 by weight and add water to make slurry. The slurry is then moulded by hand in a recycled plastic can to shape it and squeeze out water and the briquette formed is dried in the sun. Briquette production methods adopted by local communities in Kenya are described by Njenga et al. (2013b). Soil was assumed here to be non-organic, because subsoil, for example the spoil from pit digging, roadsides and riversides, is commonly used. Water is fetched from natural shallow wells and transported on foot and this process was not included.

2.1.2.5. Kerosene production and transportation. Kerosene was the other type of cooking fuel used as reference. Its production stage was considered at the refinery in Mombasa and data from the Ecoinvent database on Kerosene, at refinery/RER U were used. Electricity, hydropower, at power plant/SE U was chosen, as it is the common source of power in Kenya. Data on transportation using Lorry greater than 16t, fleet average/RER S were chosen from the Ecoinvent database. Kerosene is transported by oil companies in better and larger trucks. Kerosene is transported to Nairobi from the refinery in Mombasa, a distance of 500 km. The distance was doubled as the trucks are driven back to Mombasa empty.

2.1.2.6. Use of the three types of fuel for cooking. Real-time measurements were made on the amounts of A. mearnsii charcoal, charcoal briquettes and kerosene required in cooking the traditional meal for a standard Kenyan household of five people. The cooking efficiency tests were carried out at the Human Needs Project (HNP) open ground at Kibera slum, Nairobi, in early 2012. The amount of fuel used and time taken to cook the meal were calculated. The cooking was conducted by 23 women who lit the cooking stoves, added water and fuel as required and tasted when the food was completely cooked. In all, 850 g of charcoal briquette with calorific value of 24.5 kJ/g, 890 g of charcoal with 29.0 kJ/g, or 0.36 litres (L) of kerosene with 43 kJ/g were needed to cook the

standard meal (Njenga et al., 2014; Njenga et al., 2013b). More charcoal than charcoal briquette was required as to cook the meal as the latter burns slowly for 4 h compared to 2.5 h of the former. Cooking with charcoal briquette and charcoal was done using the commonly used improved cooking stove known as Kenya Ceramic Jiko (KCJ), while for kerosene, a cooking stove from India commonly found in Kenya was used.

Indoor air concentrations of CO, CO2 and PM2 5 were measured from burning 0.75 kg of charcoal briquettes and 0.64 kg of wood charcoal in the KCJ and 0.1 L of kerosene in the kerosene stove (Njenga et al., 2013a, 2013b). Measurements were carried out in triplicate in a kitchen measuring 3.4 m by 3 m, with a door measuring 2 m by 0.9 m and two windows each measuring 0.8 m by 0.6 m. One of the windows was kept open while the other was closed, simulating household cooking conditions. Measurements were made throughout the burning period of each type of fuel. The three measuring equipments were suspended with a rope close to each other at 1 m above and to the side of the cooking pot and stove, simulating the height of a person cooking (Fig. 3). Carbon monoxide was measured at 10-s intervals using an EL-USB-CO carbon monoxide data logger (DATAQ Instruments). Fine PM measurements were made every minute using a particulate matter meter (UCB, Berkeley Air Monitoring Group). Carbon dioxide was measured at 5-min intervals using a Taile 7001 Carbon Dioxide and Temperature meter (LASCAR). Each of the three equipment's have smoke detectors that sense real-time signals and take measurements and log the data continuously.

21.2.7. Limitations. The application of LCA to potential environmental aspects was limited to GHG emissions, quantified as GWP-ioo. The focus of the study was on the use of wood as charcoal and charcoal dust briquettes, and the production of woody biomass was not considered in detail. The implications of the assumptions on biomass production are discussed in Section 3.7. The impact analysis took into consideration renewable as well as non-renewable biomass, in order to examine the effects on GHG emissions when biomass used for energy is not replanted. CH4, NOx and N2O are important GHG from indoor cooking, they were not measured due to limitation in the number of different gases that were possible within the study and with the available equipments. These emissions however have been taken into account in wood carbonization using secondary data.

21.3. Data management and analysis

Data on indoor air concentrations of CO, CO2 and PM2.5 from cooking with charcoal briquettes, charcoal and kerosene were

Fig. 3. Measuring of emissions in a kitchen at University of Nairobi.

sourced from the emission tests carried out in a kitchen as explained earlier and the results extrapolated to the amounts of fuel used in cooking the meal. A conversion factor from measured concentrations to emission factors was calculated based on a mass balance approach, by assuming full combustion to CO and CO2 of kerosene of known chemical composition. Data on amount of wood produced was from secondary sources. Data on amount of wood produced, selected mode of transport from the Ecoinvent database and emission factors in wood carbonization, waste management of charcoal dust and cooking were entered per process and assembled at each stage of the product using SimaPro 7.3.3 software.

Data were managed using Microsoft Excel software and analysed for descriptive statistics such as mean and standard deviation using the same software. Genstart Edition 13 was used for One Way Analysis of Variance (ANOVA) (VSN International, 2012). The significance of differences between any two means was tested using the least significant difference of means (LSD). Actual probability values are presented as 'P-values' to show significance at a confidence level of 95%. The emissions and cooking tests were conducted in triplicate.

There were challenges in finding data on transport and oil refinery processes based on Kenyan situations due to the lack of LCAs performed in developing countries, so use of data from Europe was the only option readily available (Ecoinvent database). This challenge has also been encountered by scientists conducting an LCA on charcoal from sawmill residues in Tanzania (Sjolie, 2012).

CO2 taken up by biomass regrowth was calculated from the amount of C released during carbonization of wood and burning of charcoal and charcoal briquettes. This amount of C was assumed to be taken up during regrowth, in the form of CO2 from the atmosphere.

3. Results and discussion

31. Global warming potential from production and use of charcoal briquettes and charcoal

When assuming that biomass used for charcoal production would not regrow, the life cycle GHG emissions were 5.36 kg CO2 eq and 4.12 kg CO2 eq from cooking a meal with the combined charcoal and charcoal briquettes (charcoal + charcoal briquette) produced in the low and high efficiency kilns, respectively which is 1.2

Table 3

Yield of fuel and number of meals per hectare under traditional (A. drepanolobium) and improved (A. mearnsii) systems of wood production and low/high efficiency wood carbonization.

Wood production Acacia drepanolobium native Acacia mearnsii woodland plantation

Wood carbonization Low efficiency High efficiency Low High

(14%) kiln (33%) kiln efficiency efficiency (14%) kiln (33%) kiln

Yield of biomass,

t/ha & yeara Yield of charcoal,

t/ha & year Number of meals from charcoal/ha & year Additional fuel from charcoal briquettes, t/ha & year Additional meals from charcoal briquettes/ha & year

1.31 0.43 430

7.22 7.22

1.01 2.39 1011 2383

Biomass usable for charcoal production.

times lower than the charcoal alone scenario (Table 2). The charcoal + charcoal briquette include the additional meals from charcoal briquettes as presented in Table 3. This indicates potential in reducing emissions from recovering charcoal dust for briquette production. The GHG from producing energy and cooking a meal in both the charcoal + charcoal briquette and charcoal scenarios using high efficiency wood carbonization process resulted in a GWP of which was 1.3 times lower than the GWP of the low efficiency wood carbonization system (Table 2). This was due to the higher wood demand in the low efficiency system, as most of the emissions are generated by wood carbonization. The two major processes in the charcoal life cycle contributing to GWP were wood carbonization and cooking. When biomass was assumed not to be regrown, the fossil fuel reference, kerosene, resulted in a GWP 3.5 and 2.7 times lower than that of charcoal + charcoal briquettes when high kiln and low kiln were used respectively (Table 2).

If the wood used for charcoal and charcoal briquette production comes from forests that regrow after wood harvesting, the CO2 emitted will be taken up in regrowing vegetation (Table 2). In this

Table 2

Global warming potential (kg CO2 eq) from the life cycle of the charcoal briquettes, charcoal and kerosene required for cooking a standard traditional meal (githeri) in Kenya.

Charcoal + charcoal briquettes

Acacia

drepanolobium native woodland and low efficiency kiln

Acacia mearnsii plantation and high efficiency kiln

Charcoal

Acacia drepanolobium native woodland and low efficiency kiln

Acacia mearnsii plantation and high efficiency kiln

Kerosene

Wood carbonization 3.33 Refinery

Transportation 0.09

Waste management 0

Cooking 1.94

Total (non-regrowing 5.36 biomassc)

CO2 taken up by biomass -3.72 regrowth

Total (regrowing 1.64 biomassd)

0.09 0

1.94 4.12

1.00 1.53

a Not applicable. b Included in refinery. c Trees not replanted. d Trees replanted.

case, the GWP from charcoal + charcoal briquettes is reduced to an average of 1.6 kg CO2 eq per meal, which is lower than 2.0 kg CO2 eq per meal for charcoal (Table 2). Also in this case, a major source of GWP in the charcoal life cycle is CH4 from wood carbonization. This and other emissions in the life cycle of charcoal are large enough to give a higher GWP from charcoal than kerosene, even when the wood production is renewable.

Regrowing biomass is critical where low efficiency kilns are used as more CO2 is released into the air as compared to where high efficiency kilns are applied.

An LCA of eucalyptus charcoal briquettes in Brazil by Rousset et al. (2011) showed that supplying the energy content of 1 kg of briquettes resulted in 4 kg of CO2 emissions. However, details on assumptions for this emissions figure were not presented, other than that a 'sustainable eucalypt plantation' was used. In the present study, the charcoal briquettes were made from waste (charcoal dust), whereas in the Brazilian study the briquettes were the result of an industrial production process.

3.2. Land requirement for wood production

A. drepanolobium native woodland vegetation yielded 0.18 t/ha and 0.43 t/ha of charcoal per year when a low and high efficiency kiln was used, respectively. Producing wood in an A. mearnsii plantation system yielded 1.01 t/ha and 2.39 t/ha of charcoal per year when a low and high efficiency kiln was used, respectively. The higher production of wood in a plantation could be associated with the pure stand of tree species, while in native vegetation there is a mix of tree species and charcoal producers select the preferred species. However, it is important to note that this direct comparison does not account for multifunctional land use; the native vegetation in drylands is also used as grazing for livestock and a habitat for wildlife. Between 0.03 and 0.37 t/ha of additional fuel per year was produced through recycling charcoal dust for briquette production (Table 3).

As a consequence, adoption of growing of trees and shrubs on-farm for charcoal production, such as short rotation forestry, is desirable. Farm forestry, including the planting of woodlots on farms for charcoal production, would go a long way towards providing a continuous supply of raw materials (Kituyi, 2004).

3.3. Traditional versus improved wood carbonization methods

Due to its higher efficiency in converting wood into charcoal, the high efficiency kiln saved 57% of the wood used in the low efficiency kiln to produce charcoal to cook the same size and type of meal. These findings are in line with previous results on resource use in charcoal production in Kenya and Tanzania using an improved kiln compared with a traditional kiln (Pennise et al., 2001; Kituyi, 2004; Bailis, 2009; Sj0lie, 2012). This supports the on-going initiatives by organizations such as the Kenya Forestry Research Institute (KEFRI) on development and use of more efficient wood carbonization processes in the country (Oduor et al., 2006). This will yield more charcoal from each kg of wood. The data are also useful in reinforcement of regulations for the sustainable charcoal sector in Kenya, which recommends use of efficient wood carbonization processes governed by the Kenya Forest Services (Gathui et al., 2011). At the wood carbonization stage, using a high efficiency kiln reduced GHG emissions by 37% compared with the low efficiency kiln (Table 2). Thus the contribution of this stage to GWP in the product life cycle is lowered with higher efficiency in the carbonization process. With respect to GWP, the carbonization process was second to cooking in terms of environmental impacts.

3.4. Additional fuel from charcoal briquettes

The method used for wood carbonization determined the amount of wood required and consequently the area of land required for wood production. To produce enough charcoal to cook the standard meal, 3 and 7 kg, respectively, of wood were required when using the high and low efficiency carbonization process. This amount of wood included 13% that ended up as charcoal dust. Adoption of improved methods would reduce wood and energy wastage, hence lowering the land requirement and saving trees. Recovering charcoal dust for charcoal briquette production gave 0.40 and 0.03 t/ha of additional cooking fuel with the improved and traditional system, respectively (Table 3). Hence charcoal briquettes contributed 16% more fuel and 18% more meals per hectare. The higher percentage increase in meals compared with fuel yield was because charcoal briquettes were more energy efficient than charcoal in the cooking stage.

This implies that recycling charcoal dust for briquette production contributes to additional cooking fuel which reduces amount of charcoal needed. This will result into saving of trees otherwise cut down for charcoal production as well as contributing to management of urban waste. The number of meals from charcoal briquettes could be increased by carbonizing biomass unusable for charcoal production, such as branches and leaves.

3.5. Transportation, production of kerosene and waste management

Transportation of charcoal fuel to cook a standard meal from rural areas to the city resulted in 0.1 kg CO2 eq. However, transporting charcoal dust to produce charcoal briquettes accounted for zero burden, as it is collected from charcoal retail shops within the slum and transported on the back of women for distances less than 0.5 km. Packing of charcoal was excluded from the calculations because recycled bags are used. The GWP caused by transportation of charcoal could be reduced through peri-urban agroforestry, where land has less competing uses for settlement, while transporting charcoal in high capacity vehicles would be beneficial.

The kerosene production stage caused GWP of 0.48 kg CO2 eq, while its transportation to the city caused 0.05 kg CO2 eq, accounting for 31% and 3%, respectively, of the life cycle. The GWP was lower for transporting kerosene than for transporting charcoal, which could be associated with higher energy density in the former.

When charcoal dust is not recycled for briquette making, it is either burned on-site or thrown into open drains. Handling enough charcoal to cook the traditional meal resulted in 0.13 kg of charcoal dust and burning it in urban areas produced 0.3 kg CO2 eq (Table 2).

3.6. Emissions from cooking with different fuels

Using charcoal briquettes for cooking emitted the lowest amounts of CO2, CO and PM2 5 (Table 4). There was a significance difference (P < 0.05) in CO2, CO and PM2 5 between the three types of fuels. Cooking with charcoal briquettes reduced CO2, CO and PM2.5 emissions by 60%, 72% and 88%, respectively, compared with charcoal. Charcoal had the highest emission factors for CO2, CO and PM2.5 (Table 4).

At the cooking stage, using charcoal briquette reduced GWP by 40% compared with cooking with charcoal (Table 2). For charcoal briquettes, the cooking stage accounted for 100% of GWP in its life cycle, since other stages were considered to have no environmental burden as the raw material is waste in the charcoal supply chain (Table 2). For charcoal briquettes, the cooking stage was second only to wood carbonization (Table 2).

Table 4

Indoor air concentration of CO2, CO and PM2.5 and emission factors (g pollutant/meal) in household cooking. Standard deviation (SD) and least significant difference (LSD) values are given at 95% confidence level.

Type of fuel

Burning period (hour)

Amount of fuel burned

Amount of fuel per

Average indoor air concentration during cooking

Emission factor (g pollutant/ meal)

meal CO2 (ppm) CO (ppm) PM2.5 (mg/m3) CO2 CO PM2.5

Charcoal briquette 4 0.75 kg 0.85 kg 96.6 16.4 0.04 1069.7 115.8 0.2

SDa 17.3 1.7 0.03 191.7 11.8 0.1

Charcoal 2.5 0.64 kg 0.89 kg 240.6 59.1 0.37 1665.0 260.3 0.9

SD 25.2 5.4 0.2 174.1 24 0.7

Kerosene 1 0.1 l 0.361 271.9 37.7 0.3 752.8 66.3 0.5

SD 53.6 11.1 0.07 148.4 19.5 0.1

LSDb 71.1 14.4 0.2 344.3 38.2 0.9

a Standard deviation. b Least significance difference.

Cooking with kerosene caused a GWP of 1.0 kg CO2 eq, accounting for 65% of its life cycle emissions. Cooking with kerosene caused a 2fold and 1.3-fold reduction in GWP compared with charcoal and charcoal briquettes, respectively, at the cooking stage. However, cooking the traditional meal with kerosene cost 45 Ksh (US$0.6) which is 2-fold and 6-fold higher than charcoal (Ksh 26 and US$0.35) and charcoal briquettes (Ksh3 and US$0.04), respectively (Njenga et al., 2013a). Cooking with kerosene also requires cooking appliances that the poor might not be able to afford. The calorific value of charcoal briquettes and charcoal used in the cooking tests was 24.5 kJ/g and 29.0 kJ/g, respectively (Njenga et al., forthcoming), while that of kerosene was 43 kJ/g. Despite the calorific values of these fuel types, the time taken to cook the traditional meal, which was continuously tasted by 23 women to establish when it was ready, took 178 min, 168 min and 166 min for charcoal briquettes, charcoal and kerosene, respectively. The only difference was that a low amount of kerosene was used (Table 4), which may also have contributed to its low emissions for the whole cooking period.

Cooking is also a stage where fuel efficiency and emissions may be influenced by the type of stove used (Edwards et al., 2003). For instance, use of the KCJ as opposed to the traditional type for cooking reduces emissions of CO by 15% (Kituyi et al., 2001). Hence fuel efficiency and emissions, and consequently the life cycle GWP, may differ from one household to another depending on the type of cooking stove used. To address the GWP caused by cooking, there is a need for urgent technological and policy interventions to reduce emissions at this stage, such as introduction of efficient cooking stoves. There is also a need to assess the health implications of cooking with the three fuels studied.

3.7. Carbon dioxide (CO2) uptake by biomass regrowth

The wood carbonization and cooking stages were the main processes contributing to GWP from the charcoal briquette and charcoal life cycles, but a large part of GWP was in the form of CO2 from biomass combustion. Hence uptake of CO2 by regrowing biomass significantly lowers the total enviromental burden from these two types of fuel by cancelling out most of the emissions (Table 2). This emphasises the importance of replanting trees cut down for charcoal, use of mature stems leaving others to grow from the same tree and choosing tree species that coppice well and those that naturally regenerate. It is equally critical to recommend suitable species in different agroclimate conditions. For instance, the studied A. mearnsii grows well in Agroclimate Zones I—III (Maundu and Tengnas, 2005). It is suitable for woodlots but should not be intercropped, as it competes for nutrients, and should be well managed, as it is potentially a weed. It is also important to ensure that the tree species recommended do well in naturally dry conditions, such as Acacia tortilis for Agroclimate Zones (IV—VII) (Maundu and Tengnas, 2005).

Recent developments in determining the GHG balance of bio-energy systems have shown that assumptions about forest carbon cycles have large impacts on the resulting GHGs (Helin et al., 2013). As the present study focused on later stages in the life cycle, not on primary production of biomass, the forest carbon cycles were not described or analysed in detail. However, our systems limitations and assumptions contain some implicit assumptions about carbon cycles. In the charcoal and charcoal briquette scenarios, it was assumed that there were no changes in soil C or vegetation C. Only CO2 uptake from harvested biomass was included, which is an implicit assumption that there is no net uptake or loss of C in soils or vegetation in these forest systems. This is a simplistic assumption, and better knowledge about C cycling in forest ecosystems and plantations in Kenya could shed light on this issue and show if other assumptions would better describe the situation (Bailis, 2009). Furthermore, in the fossil fuel reference case, with kerosene used as fuel, no assumption was made on CO2 uptake or emissions from land use. This is equivalent to assuming no change in C stock due to land use, i.e. no net growth of forest if it were not used for supply of wood for charcoal production. This could be the case in a mature forest or a degraded forest with no re-growth. An alternative assumption could have been that if kerosene were to be used instead of charcoal, there would be a regeneration of biomass in forests, i.e. a net uptake of CO2. That would give the kerosene system a lower GWP than with the assumptions used in this study.

The carbon-neutral forest systems described included a regeneration period of 9 years for A. mearnsii plantation and 14 years for A. drepanolobium. During that period, CO2 emitted by biomass is gradually taken up by regrowing biomass, but the CO2 spends some time in the atmosphere and contributes to GWP during that time. This has not been included in previous LCAs, but its omission has been questioned in recent years and different methods to account for CO2 dynamics have been suggested (Helin et al., 2013). To explicitly include the temporal aspect of GHG emissions caused by biomass growth and harvest, a time-dependent indicator is required (Ericsson et al., 2013). A single-value indicator such as GWP cannot fully describe the time-dependent climate of GHG emissions and uptake. A simple method for accounting for the climate impact of biogenic CO2 is the GWP bioindex proposed by Cherubini et al. (2011).

4. Conclusions and recommendations

Recovering charcoal dust for charcoal briquette production contributes additional fuel, which improves energy security and may reduce deforestation. It also has potential in reducing GHG as it produces low emissions from the additional fuel produced. In the life cycle of charcoal + charcoal briquettes and charcoal, the highest GHG emissions were from wood carbonization and cooking. Although kerosene had lower GWP than charcoal, most of its GHG

emissions are inevitable as it is a fossil fuel, whereas the GWP of charcoal can be reduced by adopting better practices in wood supply, wood carbonization and household use. For sustainability in supply of charcoal briquettes and charcoal as cooking fuels in developing countries, it is crucial to replant trees cut down for charcoal or to select tree species that regenerate naturally and/or that coppice well, to ensure the presence of biomass for uptake of carbon dioxide, hence maintaining the neutral impact of biomass energy on GWP.

Acknowledgements

This work was made possible with financial support from the International Development Research Centre (IDRC), World Agro-forestry Centre (ICRAF), African Women in Agricultural Research and Development (AWARD) and the Borlaug LEAP. The technical support of RaidaJirjis, Fridah Mugo, Nelly Oduor, Joshua Cheboiwo and Jan Vandenabeele is highly appreciated. The support of Product Ecology Consultants in provision of a free SimaPro Faculty licence and technical advice is gratefully acknowledged.

References

Afrane, G., Ntiamoah, A., 2011. Comparative life cycle assessment of charcoal, biogas, and liquefied petroleum gas as cooking fuels in Ghana. J. Ind. Ecol. 15, 539—549. Bailis, R., 2009. Modeling climate change mitigation from alternative methods of

charcoal production in Kenya. Biomass Bioenerg. 33,1491—1502. Cheboiwo, J., Mugo, F., January-March 2011. Black wattle for Fuelwood. Acacia mearnsii offers potential for profits while meeting the deficit for charcoal and firewood. Miti. Tree Bus. Mag. Afr. (9), 28—29. www.betterglobeforestry.com/ miti-magazine.

Cherubini, F., Peters, G., Berntsen, T., Stromman, A., Hertwich, E., 2011. CO2 emissions from biomass combustion for bioenergy: atmospheric decay and contribution to global warming. Glob. Change Biol. Bioenerg. 3 (5), 413—426. Ecoinvent Centre, 2010. Ecoinvent Data v2.2 Ecoinvent Reports n. 1—25. Swiss

Centre for Life Cycle Inventories, Dubendorf, Switzerland. Edwards, R.D., Smith, K.R., Zhang, J., Ma, Y., 2003. Models to predict emissions of health-damaging pollutants and global warming contributions of residential fuel/stove combinations in China. Chemosphere 50, 201—215. Ericsson, N., Porso, C., Ahlgren, S., Nordberg, A., Sundberg, C., Hansson, P.-A., 2013. Time dependent climate impact of a bioenergy system - methodology development and application to Swedish conditions. Glob. Change Biol. Bioenerg. vol. 5 (5), p. 580—590. Gathui, T., Mugo, F., Ngugi, W., Wanjiru, H., Kamau, S., 2011. The Kenya Policy Handbook. In: Current Regulations for a Sustainable Charcoal Sector. Practical Action Consulting East Africa. Helin, T., Sokka, L., Soimakallio, S., Pingoud, K., Pajula, T., 2013. Approaches for inclusion of forest carbon cycle in life cycle assessment — a review. Glob. Change Biol. Bioenergy 5, 475—486. IEA, World Energy Outlook, 2006. IEA/OECD, Paris, France, p. 596. IPCC, 2007. Climate change 2007: the physical science basis. In: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge and New York, NY, p. 212.

Kakuzi, 2003. Economics of Charcoal Production Using the Kakuzi Charcoal Kilns. Thika, Kenya.

Karekezi, S., 2002. Renewables in Africa—meeting the energy needs of the poor. Energy Policy 30, 1059-1069.

Kaygusuz, K., 2011. Energy, electricity and socioeconomic transformation of rural regions in developing countries. Energy Sources B Econ. Plan. Theory 6 (4), 384-394.

Kenya Government, 2010. 2009 Kenya Population and Housing Census. In: Population Distribution by Administrative Units, vol. 1 A. Government Printers, Nairobi.

Kituyi, E., Marufu, L., Wandiga, S., Jumba, I., Andreae, M., Helas, G., 2001. Carbon monoxide and nitric oxide from biofuel fires in Kenya. Energ. Conv. Manag. 42 (13), 1517-1542.

Kituyi, E., 2004. Towards sustainable production and use of charcoal in Kenya: exploring the potential in life cycle management approach. J. Clean. Prod. 12, 1047-1057.

Maundu, P., Tengnas, T. (Eds.), 2005. Useful Trees and Shrubs for Kenya. Technical Handbook No. 35. World Agroforestry Centre Eastern and Central Africa Regional Programme (ICRAF-ECA), Nairobi, Kenya.

Ministry of Energy (MoE), 2002. Republic of KENYA Study on Kenya's Energy Demand, Supply and Policy Strategy for Households, Small Scale Industries and Service Establishments. Project Report.

Mugo, F., Nungo, R., Odongo, F., Chavangi, N., Abaru, M., 2007. An assessment of the energy saving potential and heat loss pattern in fireless cooking for selected commonly foods in Kenya. In: CARPA Working Paper Series, No. 2. CAPRA.

Mugo, F., Poulstrup, E., 2003. Assessment of Potential Approaches to Charcoal as a Sustainable Source of Income in the Arid and Semi-arid lands of Kenya, DANIDA and RELMA: Final Report No. 72.

Mutimba, S., Barasa, M., 2005. National Charcoal Survey: Summary Report. Exploring the Potential for a Sustainable Charcoal Industry in Kenya. Energy for Sustainable Development Africa (ESDA).

Njenga, M., Yonemitsu, A., Karanja, N., liyama, M., Kithinji, J., Dubbeling, M., Sundberg, C., Jamnadass, R., 2013a. Implications of charcoal briquette produced by local communities on livelihoods and environment in Nairobi, Kenya. lnt. J. Renew. Energ. Dev. (IJRED) 2 (1), 19-29.

Njenga, M., Karanja, N., Jamnadass, R., Kithinji, J., Sundberg, C., Jirjis, R., 2013b. Quality of briquettes produced locally from charcoal dust and sawdust in Kenya. J. Biobased Mater. Bioenergy 7 (3), 315-322.

Njenga, M., Karanja, N., Kithinji, J., liyama, M., Jamnadass, R., 2014. Combustion and emissions characteristics of fuel briquettes made from charcoal dusts and sawdust of selected tree species in Kenya. Energy for Sustainable Development, (forthcoming).

Oduor, N., Githiomi, J., Chikamai, B., 2006. Charcoal Production Using lmproved Earth, Portable Metal, Drum and Casamance Kilns. Kenya Forestry Research lnstitute (KEFRl).

Okello, B.D., O'Connor, T.G., Young, T.P., 2001. Growth, biomass estimates, and charcoal production of Acacia drepanolobium in Laikipia, Kenya. For. Ecol. Manag. 142,143-153.

Olz, S., Sims, R., Kirchne, N., 2007. Contribution of Renewables to Energy Security. lnternational Energy Agency.

Pennise, D., Smith, K., Kithinji, J., Rezende, M., Raad, T., Zhang, J., Fan, C., 2001. Emissions of greenhouse gases and other airborne pollutants from charcoal making in Kenya and Brazil. J. Geophys. Res. 106 (20), 143-24.155.

Rousset, P., Caldeira-Pires, A., Sablowski, A., Rodrigues, T., 2011. LCA of eucalyptus wood charcoal briquettes. J. Clean. Prod. 19 (14), 1647-1653.

Sj0lie, H.K., 2012. Reducing greenhouse gas emissions from households and industry by the use of charcoal from sawmill residues in Tanzania. J. Clean. Prod. 27,109-117.

SimaPro 7.3.3, 2011. Pré Consultants BV. Amersfoort, The Netherlands.

Terra Nuova, AMREF, EC Briquette Technology, 2007. An lnformative Research Report on Briquette Making Technology in Kenya. Project Report.

VSN lnternational, the Numerical Algorithms Group, 2012. GenStat Version Statistical Package for Windows.