90 Table 5.7: Sensitivity analysis of climatic, edaphic, topographical and land use factors affecting the raw material production potential. 127 Table 8.6: Suitability score when the minimum relative humidity of each month is ideally suitable for sorghum cultivation.
INTRODUCTION
- The Need for Biofuel Production
- The Importance of Biofuel in South Africa
- The Impact of Biofuel Production
- The Problem Statement
- Research Questions
- Purpose of the Study
- Assumptions/Limitations
- Chapter Overview
A detailed assessment of areas suitable for commodity production is required to determine whether such areas exist, particularly in the former homelands. The main research question addressed in this study is: Which areas in South Africa are optimally and suboptimally suitable for the production (ie cultivation) of soybeans, sorghum and sugar cane.
PRODUCTION OF BIOFUELS
Definition of Biofuels
- Classification of biofuels
- First and second generation feedstocks
These advanced conversion technologies are still in development and are not currently considered economically viable (Ravindranath et al., 2011). Since bioethanol has a high octane number, it can therefore be used to replace octane boosters, increasing the efficiency of combustion engines (John et al., 2011).
Overview of the Emerging Biofuels Industry in South Africa
- The national biofuels strategy
- Mandatory blending rates
- Proposed biofuel plants
A “granted” license status means that the applicant has not met all the requirements (DoE, 2014). Issued means that the applicant has met all requirements and is authorized to manufacture.
Strategic Biofuel Feedstocks
- Soybean (Glycine maximum)
- Grain sorghum (Sorghum bicolor)
- Sugarcane (Saccharum officinarum L.)
According to Spark et al. 2010), the use of by-products of biofuel processing can contribute significantly to the economy of South Africa. In addition, there is ongoing research in South Africa into sugarcane varieties suitable for energy production (Jewitt et al., 2009a).
Other Potential Feedstocks
- Sugarbeet (Beta vulgaris)
- Sweet sorghum (Sorghum bicolor L. Moench)
- Cassava (Manihot esculenta crantz)
- Maize (Zea mays L.)
- Sunflower (Helianthus annuus L.)
- Canola/Rapeseed (Brassica napus L.)
- Jatropha (Jatropha curcas)
- Moringa (Moringa oleifera Lam.)
The seasonal precipitation range considered sufficient for sweet sorghum is mm in the growing season. It is believed that Jatrofa was spread by the Portuguese navigators and is now also found in the tropics and subtropics (Brittaine and Lutaladio, 2010).
MAPPING SUITABLE PRODUCTION AREAS
Factors Affecting Plant Growth and Distribution
- Rainfall
- Temperature
- Relative humidity
- Soils and topography
- Land use
Schulze and Maharaj (2007d) highlighted that plant growth is affected by the duration of the rainy season, i.e. Additionally, slope (i.e., the rate of change of elevation over distance) is also commonly considered a static feature of the physical landscape.
Overview of Land Suitability Assessment
- Land suitability assessment
- Climatic constraints
- Technical constraints
- Environmental considerations
- Combining individual suitability ratings
The importance of soil depth and slope on plant growth and plant distribution was discussed earlier in Section 3.1.4. According to Camp (1995; cited by Sibanda, 2008), evaluation and investigation of soil factors (eg soil depth and slope) should be an important criterion in soil evaluation studies. This case study is further discussed in Section 3.3.1 along with three other GIS-based studies to assess land considered suitable for biofuel feedstock production.
GIS-based Case Studies
- Potential of Jatropha curcas production in South Africa
- Physical potential of bioethanol processing plants in Kenya
- The biofuels scoping study for South Africa
- Bioenergy production in semi-arid and arid areas in sub-Saharan
- Summary and the way forward
The scoping study on crop/tree water use for biofuels in South Africa (Jewitt et al., 2009a) identified potential growth areas for selected biofuel feedstocks. Climatically suitable areas for optimal growth of selected raw materials are highlighted in dark green in Figures 3.5 to 3.7 (Jewitt et al., 2009a). For Figure 3.5, the mapping criteria used for soybean were based on a seasonal rainfall of 550 – 700 mm and monthly mean daily mean temperature of 20 – 30 °C.
METHODOLOGY
Data Sources
- Rainfall
- Temperature
- Relative humidity
- Slope
- Soil depth
- Land use
- Summary
According to Schulze et al. 2007b), the uncorrected true vapor pressure is predictable from month to month in South Africa using mainly geographic factors and regression equations. When calculating the daily RHmin and RHmax, it was assumed that the uncorrected actual vapor pressure is constant throughout the day. 2007b) Land use Land use in South Africa BGIS Bhengu et al. 2008) Protected areas Formal and informal protected areas v.
Growth Criteria for Each Feedstock
Mapping Software and Technology
Data Manipulation and Analysis
These latitude/longitude grid cells then represent non-square cells of ~200 by 231 m, which corresponds to a cell size of 0.0020833. The analysis generated new datasets from existing raster data to create different criteria that served as input to the land suitability assessment. The Raster Calculator in Spatial Analyst was also used to overlay the input criteria and complete the land suitability assessment.
Elimination of Unsuitable Areas Based on Threshold Limits
Raster grid cells are always square in the projected coordinate system, but the analysis was performed in the geographic coordinate system (cape datum), with reported cell size units in degrees latitude/longitude. This exercise was repeated for other raster climate datasets (eg temperature and relative humidity). These new data layers then formed the input criteria for assessing the suitability of land for growing biofuel feedstocks.
Land Suitability Evaluation Procedure and Ranking of Suitability Criteria
- Rainfall
- Temperature
- Relative humidity
- Soil depth
- Slope
- Overall weighting of each suitability criteria
The final thresholds used to derive seasonal rainfall ranges for each FAO-based suitability class are given in Table 4.5 for soybean. These subjective weightings were based on expert opinion (Bertling and Odindo, 2013) and ranged from most important (40 %) to least important (10 %), as shown in Table 4.11. Each goodness-of-fit score was calculated by multiplying the reclassified rank (i.e. data) by the decimal weight.
GIS Approach
- Distribution of seasonal rainfall over the growing season
- Weighting of monthly temperatures
- Weighting of monthly relative humidity
- Weighting of soil depth and slope
- Normalisation of overall suitability score
- Masking with land use
The Kc values were again used to weight each month's ranking to produce a fitness score for each month, which was then summed (Table 4.17). If the monthly temperature is within the ideal range for each of the five months during the growing season (eg the rank of 3 assigned to each month), it produces a maximum temperature suitability score of 0.6 out of 3 (Table 4.19). If the moisture level is within the ideal range for each of the five months during the growing season (eg the rank of 3 assigned to each month), it produces a maximum moisture suitability score of 0.3 out of 3 (Table 4.20).
Summary
Functional “no-go” areas refer to land uses that are not currently considered suitable for the cultivation of commodities (see section 3.2.1) and include commercial forest plantations and natural and degraded lands. All grid cells found suitable for commodity cultivation (S1, S2 or S3), but overlapping with land use areas classified as N1 or N2, were excluded (or filtered out) using GIS. Taking current land use into account therefore reduces the total amount of arable land available for growing raw materials.
RESULTS AND DISCUSSION
Elimination of Unsuitable Areas
Similarly, a comparison of figure 8.1 (appendix) with figure 3.6 from the scoping study shows a significant increase in suitable area for sorghum. The reason for this is that the scoping study mapped optimal growth areas, whereas this study mapped suitable growth areas (which range from optimal to marginal). Finally, the difference between Figures 8.2 and 3.7 for sugarcane is explained by the use of annual temperature in the scoping study and monthly temperatures in the present study.
Land Suitability Evaluation
- Rainfall
- Temperature
- Relative humidity
- Slope suitability
- Soil depth
- Summary
- Land use
- Overall land suitability score
- Finalising the map legend
- Sensitivity analysis
The weighting of 0.2 in Figure 5.2 refers to Table 4.11, values closer to 0 are unsuitable and values approaching 0.2 are highly suitable. The weighting of 0.1 in Figure 5.3 refers to Table 4.11, values closer to 0 are unsuitable and values approaching 0.1 are highly suitable. The weighting of 0.2 in Figure 5.4 refers to Table 4.11, values closer to 0 are unsuitable and values approaching 0.2 are highly suitable.
Sensitivity to Crop Coefficient Values
- Basal vs. single crop coefficient approach
- International vs. local K c approach
However, the use of local Kc values slightly reduced the area considered suitable for soybean, sorghum and sugarcane production, compared to FAO's Kc approach. This study showed that Kc from international crops should not be used, but that local Kc should be applied. According to Allen et al. 1998), the user is strongly encouraged to obtain appropriate local information (crop coefficients).
Map Validation
- Comparison with estimated yields
- Comparison with observed yields
The final map (Figure 5.12) included an overlay of soybean yields obtained from Stats SA (Stats SA, 2007) for dry land conditions and aggregated to magisterial district level. Dry land yields were used to identify lower production areas (< 1.8 t.ha). 1) and higher production areas (> 2.0 t.ha-1). It is unfortunate that the yield data is aggregated to the magisterial district level and not made available at the company level. A similar exercise was undertaken to determine whether marginal areas “match” with lower yielding areas, with good agreement in the Free State and Mpumalanga Province.
Location of Biofuel Processing Plants
However, the overlay proved useful in assessing whether highly suitable areas corresponded to areas of higher productivity, with some agreement in Kwa-Zulu Natal near the border with Lesotho. It is likely that only a few farms occur in this region and a particular drought-resistant cultivar or variety is grown. It was not possible to compare the sorghum suitability map with yield data obtained from the Stats SA census.
Weighted vs. Boolean Mapping Approach
In addition, the maps show that KwaZulu-Natal is the only province capable of producing all three feedstocks in sufficient quantities to meet the expected biofuel demand as determined by the mandatory blending rates. A comparison of the total area suitable for soybean and sorghum production highlights that the elimination approach shows more than double that highlighted by crop coefficient approaches. In other words, given that the local Kc approach provides the most realistic estimates of land suitable for commodity production, these areas are approximately half of those produced by the Boolean approach.
Summary
Based on the figures in Table 5.8, soybeans show the greatest potential for agricultural expansion in South Africa, while sugarcane has the least expansion potential.
CONCLUSIONS
Summary of Approach
Summary of Findings
This study found that the Boolean mapping methods (ie where areas are classified as suitable or unsuitable) tend to overestimate the land area considered suitable for commodity growth. The importance of rainfall distribution across the growing season and not just total seasonal rainfall was also highlighted in this study. The importance of using crop coefficients applicable to local growing conditions (and not obtained from overseas studies) was also highlighted.
Recommendations for Future Research
A unique methodology, based on the concept of crop coefficients, was used to determine when the crop needs the most rain during its growing cycle. The research subsequently showed that basic crop coefficients (based only on transpiration) should not be used, but that coefficients for individual crops, which also take into account groundwater evaporation, should be part of the approach. The results of this study should help pave the way for the emerging biofuels industry by providing information at a more appropriate level to decision makers and local governments.
Adaptation of cassava (Manihot esculenta) to the arid environments of Limpopo, South Africa: growth, yield and yield components. An economic evaluation of soybean-based biodiesel production on commercial farms in Kwazulu-Natal, South Africa. The development of a hydrologically enhanced Digital Elevation Model and derived products for South Africa based on the SRTM DEM.
APPENDIX
Grain sorghum
If the monthly temperature is within the ideal range for each of the five months during the growing season (eg the rank of 3 assigned to each month), it produces a maximum temperature suitability score of 0.6 out of 3 (Table 8.4). . Basically, a total of five new reclassified temperature grids were generated, then weighted and summed to calculate the temperature suitability score. If the moisture level is within the ideal range for each of the five months during the growing season (ie the rank of 3 assigned to each month), it produces a total suitability score of 0.3 out of 3 (Table 8.6).
Sugarcane
The soil depth thresholds were also derived from the literature review and are deeper than those for soybeans, as grain sorghum is a deeper root crop (which also helps prevent lodging). In addition, this study also took into account the crop coefficients for ratooned sugarcane from FAO (2013), as shown in Table 8.9. These thresholds are deeper than those for grain sorghum, as sugarcane is not as drought tolerant.
