Carbon Footprint
A key advantage of using SugarCaneModel for decision-making is that the many interdependencies involved in any project, large or small, can be included holistically in a single model, and the potential outcomes can be judged transparently and fairly against a single measure. This single measure is often financial: it may be annual profit, payback, NPV, or any other parameter which is deemed appropriate. Financial measures are extremely useful as almost all the significant impacts of a project can be measured on a monetary basis.
Just as the financial impact can be captured well by a single measure, the impact on climate change can also be measured transparently and fairly, via the estimation of overall greenhouse gas (GHG) emissions, or “carbon footprint”.
SugarCaneModel includes carbon footprint in its calculations and outputs. The method used is generally as published and applied by Bonsucro in accrediting sugar producers for the Bonsucro Certification System, the first global metric standard for sugar cane. The Bonsucro method accounts for direct and indirect energy use and GHG emissions in the following areas:
- Agriculture (irrigation, chemical use, cane burning, transport fuel use, field residues)
- Fossil fuels burnt
- Electricity imports/exports (exports result in a credit)
- Process chemicals used
- Allocation to co-products (e.g. molasses, ethanol)
The example below illustrates how SugarCaneModel can be used to estimate the potential variation in GHG emissions and highlight the key factors driving those variations. A special simulation was run of a sugar plantation and factory supplying raw sugar into the Middle East, in which a large number of the input parameters were allowed to vary widely. The intention was not to reflect an actual single scenario; rather, it was intended to reflect the diversity of potential scenarios worldwide. Examples of input parameters that were allowed to vary during the simulation include:
- Location of factory (varying between India, Brazil and the Philippines)
- Cane quality (fibre, sucrose & purity)
- Cane yield
- Irrigation water usage
- % of irrigation pumps which are diesel or electric
- Amount of fertilisers, herbicides and insecticides used
- Extent of cane burning
- Extent of mechanisation of planting & harvest
- Amount of trash recovered from the fields and used in the boiler
- Mill extraction, imbibition water usage and bagasse moisture
- % of mill drives electrified
- Evaporated syrup brix
- Process energy efficiency
- % of molasses processed into ethanol
- Destination of filter mud and distillery vinasse (returned to cane fields, sold as product or disposed of via landfill)
- Whether power can be exported to the grid
- Whether condensing turbines (in addition to backpressure type) were included
- Boiler pressure
- Transport distances (e.g. field to mill, mill to port)
The chart below shows the probability distribution of the total GHG emissions (measured as g CO2eq (equivalent) per kg sugar delivered, i.e. including sea freight), resulting from 2000 simulations of the model.
As might be expected, given the wide variability assigned to the inputs, there is a wide variability in the estimated GHG emissions. The mean value is around 610 g CO2eq per kg sugar, and the 90% confidence interval is around 255 to 820 g/kg. It is also interesting to note that in a small number of cases (~1%), the GHG emissions are negative.
The chart below shows the sensitivity analysis chart, indicating the top ten drivers behind the variability in emissions. The values shown reflect the indicative swing caused by variations in the input parameter. For example, variations in boiler pressure caused an average swing in total GHG emissions of around 70 g/kg. It is important to remember here the nature of this ‘special’ simulation, in which large numbers of variables are allowed to vary widely. In a simulation of a specific case, the impact of boiler pressure could be much greater (for example, if power is exported to the grid) or much lower (if no exports are possible).
The purpose of this ‘special’ simulation is to take a broad look at which factors drive the carbon footprint up or down, and the results show that these factors can be broadly divided into two categories:
(a) Power generation. The single biggest factor is whether or not power is exported to the grid. Other factors include: country of origin; whether condensing turbines are used; boiler pressure; electrification of mill drives; cane fibre content; and turbine efficiency.
(b) Agriculture. Important factors here are the amount of nitrogen applied in cane growing, the cane yield achieved and the irrigation water usage.
One interesting driver is the country of origin. This has two effects on the GHG emissions. Firstly it affects the credit gained for power export, as emissions displaced by the power exported varies by country according to the efficiency of their power generation facilities and nature of their fossil fuel usage (as per the Bonsucro standard). Secondly (and to a lesser extent), it affects the emissions due to sea transport.
Having identified (a) that in some cases the emissions level is negative, and (b) the key drivers, it is of interest to investigate how low the emissions level can be driven. A second simulation was run in which the key parameters were fixed to represent a low-emissions scenario. Reflecting the key drivers identified above, the parameters fixed were:
- Country of origin: India (of the three countries simulated, India has the highest emissions credit for power export and the shortest sea transport distance)
- Nitrogen usage: 50 kg/ha
- High pressure boiler (100 barg) with condensing turbines and export of surplus power
- Cane yield: 100 t/h
- Mill drives electrification: 100%
The results are shown below. The emissions (in terms of absolute metric tonnes of carbon dioxide equivalent per year, and then converted into g CO2eq per kg sugar produced) are categorised in terms of agriculture, cane transport, processing (excluding power exports), local product transport (i.e. to local market or to port). A credit for power export is then applied to obtain the net total emissions value. This total value is then allocated to all products according to their relative economic values. Finally, the emissions due to raw sugar transport (sea freight to the Middle East) are added to the sugar allocation to arrive at the net g CO2eq per kg sugar produced.
|
|
mt CO2eq/y |
g CO2eq per kg sugar |
|
Agriculture |
59,466 |
281 |
|
Cane transport |
6,058 |
29 |
|
Processing |
24,416 |
115 |
|
Local product transport |
3,145 |
15 |
|
Credit for power export |
-272,799 |
-1,290 |
|
Total |
-179,714 |
-850 |
|
Allocation to molasses |
-8,036 |
-38 |
|
Allocation to ethanol |
-11,543 |
-55 |
|
Allocation to sugar |
-160,136 |
-757 |
|
Raw sugar transport |
4,888 |
23 |
|
Total |
-155,248 |
-734 |
Two conclusions immediately jump to mind from these results. Firstly, it is possible for sugar to be produced (and transported to a port refiner) at a high negative carbon footprint. At around -730 g CO2eq per kg of sugar, the carbon footprint estimated in the above case is providing more emissions savings than is actually generated in the mean case (around 610 g CO2eq) from original simulation.
Secondly, such high negative carbon footprints are only possible due to the high credit allocated to power export. In this case, around 33MW of power is exported, and as the factory is located in India, this power export is converted to an emissions credit using a factor of 0.253 kg CO2/MJ (according to the Bonsucro standard). The equivalent factor in Brazil is more than ten times lower, at 0.022 kg CO2/MJ, reflecting the different fuel mixes used in each country and the relative generation efficiencies. Therefore the same factory based in Brazil would be allocated an emissions credit of around 110 g CO2eq/kg (compared to 1290), and hence the overall carbon footprint would actually be positive, at around 450 g CO2eq/kg.
This is an illustration of how SugarCaneModel can be used to (a) estimate the carbon footprint associated with any scenario, (b) provide the confidence ranges for those estimates, and (c) analyse the key factors affecting the carbon footprint.
