The costing of machinery operations has often been a problem in economic analysis of experimental results. In Mallee work, the solution has been to refer to Neville Hall’s Mallee Gross Margins (Hall 2001), or to use local contract rates (eg. BCG). Neville’s numbers are becoming somewhat dated, as machinery width, speed and value have changed since the book was published. Farm sizes have also changed, which has altered the economies of scale. Nonetheless they are still a much better guide than nothing, as are local contract rates. When comparing between farming systems, rather than assessing different options within a system, the issues around machinery costing become more important and complex.
An initial attempt to analyse machinery needs at the BCG systems site was made in early 2009 (Jones 2009), but did not express costs in terms that could be compared either with contract rates, or the Mallee Gross Margins. Doing so is the aim of this paper.
Work rate
The systems have similar work rates for sowing and tillage, and generally for spraying (Figure 1). The spraying work rate for the Hungry Sheep system is slower, because that system uses a narrower boom (24m). Note that these work rates do not include an efficiency factor (eg. filling spray tanks or seeders, shifting machinery, fixing things, overlap, headlands). The work rates are slightly slower for spraying than estimated in Hall (2001). Hall used a wider boom (34 vs 30.5m), travelling faster (16 vs 15 km/hr). The work rate for sowing and tillage is about 50% slower in Hall, a function of wider machinery (12m vs Hall’s 9m) travelling faster (10-12 km/hr vs 9 km/hr).
Figure 1. Work rate of operations in the different systems, and listed in Hall (2001).
Variable costs
Taking a broad interpretation, the variable costs of machinery also include the hour- and hectare-related elements of depreciation. Across the systems in the systems trial, these work out to be approximately twice the fuel cost (Figure 2). Note that hectare-related depreciation for implements estimated here relies on some tenuous assumptions; more sophisticated depreciation models (eg. the Tozer 2005 approach used with tractors and harvesters) are not yet available for implements.
Variable costs are higher for seeding in the No Till versions of systems, because seeders do few hectares per year, and it has been impossible to separate the hectare- and year-related elements of depreciation. A similar effect (fewer hectares tilled per year) makes tillage costs higher in the Reduced Till system. Sowing is most expensive in the No Till and Reduced Till systems because it has been assumed that new no-till bars are used. A second-hand, chisel-plough conversion (as used in Fuel Burner and Hungry Sheep No Till) is slightly less expensive.
The variable costs for spraying are low across all the systems (slightly higher for Hungry Sheep, because of the narrower boom), and the difference between tillage and spraying approximately $4-6/ha. If labour is valued at $40/hr (eg. including all on-costs, and allowing for inefficiencies), this adds between $1/ha (most spraying operations) to $3.30/ha (most seeding/tillage). The net difference would be between $6-8.30/ha.
Compared to the Hall (2001) costings, the components of the spraying and tillage/sowing costs per hectare are quite similar. Increasing fuel costs ($0.35/l to $1.00/l) have offset improvements in fuel efficiency assumed in these estimates.
Figure 2. Variable costs of operations in the different systems.
Harvest
Harvest costs differed mainly according to header used, with slightly higher costs for the lower capacity header used by the Fuel burner, Hungry sheep and No till systems (Figure 3). At lower yields, harvest rate was limited mainly by travel speed and front width, leading to quite similar harvest costs. At higher yields, threshing capacity became the limiting factor and harvest costs were $5/ha cheaper for the Reduced till system. The yield point at which the switch between speed- and capacity-limited harvest rate occurs is about 1.5 t/ha. Despite the relationship being bi-linear, a linear fit to the data was a good approximation, with the 18.4$/t/ha slope implying an increase of header costs of $8.4/t/ha, once the $10/t cartage fee was deducted.
Figure 3. The variable cost of harvest and cartage in relation to yield, for all crops at the systems site. The cartage cost was $10/t.
Fixed costs
The fixed costs of machines includes age-related and driveaway (for new machines) components of depreciation, and foregone interest on salvaged capital. These costs do not include finance, registration, shelter or insurance.
Fixed costs on a farm basis were highest for No Till and Reduced Till systems (Table 1), due mainly to the high annual cost of 2cm GPS guidance and new seeding bar chosen for these systems. All no till systems benefit from not having the range of machinery normally associated with tillage. This led to the no till version of Fuel Burner and Hungry Sheep systems having the lowest total fixed cost.
When considering single paddocks, it is useful to have a per-hectare measure of fixed costs, and the most sensible way of considering cropping costs seems to be to relate them to average cropped area. On this basis, the Hungry Sheep No Till system is least expensive (Table 1). Adjusting for cropped area brings the No Till, Reduced Till and Fuel Burner fixed costs to a similar level.
These assessments do not include machinery assumed common to all systems – for example truck, augers, storage, field bins – which may in practice also make big differences between farms.
Table 1. Fixed costs for machines used on a hypothetical 2000ha farm in the BCG Farming Systems trial.
Contract rates
Previous analyses of the systems trial have used 75% of a local ‘contract rate’ for operations. The 75% factor was intended to remove the 25% profit margin assumed by Hall over and above real cost + labour, but probably should have been 80% (the 25% would have been calculated on real cost + labour, not the final contract rate). Compared to the 75% contract rates used (Table 2), the total variable costs (including labour at $40/hr) for the Fuel Burner system are quite a bit lower. In theory, the difference should be made up by the average fixed cost per hectare.
The real costs in Table 1 have not been attributed to each hectare (or had the tractors split according to operation), but a comparison can be made by estimating the number of hectares worked by each machine and multiplying the difference in cost to get a total implied fixed cost for the 75% contract rates (last column of Table 2). The total is quite a bit higher than the total in Table 1 (which includes a harvester and guidance), and suggests that the cost of tillage and particularly sowing was probably over-estimated.
Table 2. The 75% contract rates used in previous Systems trial analyses, compared to the variable costs (including labour) estimated for the Fuel burner system.
Discussion
References
Hall N (2001) 'Mallee Gross Margins 2001 - 2002.' (Department of Natural Resources and Environment: Swan Hill).
Jones BR (2009) Farming Systems Trial Machinery. BCG 2008 Season Research Results - Members Only Edition.
Tozer PR (2005) Depreciation Rates for Australian Tractors and Headers - Is Machinery Depreciation a Fixed or Variable Cost? In 'Australian Agricultural and Resource Economics Society Conference 2005'. (Australian Agricultural and Resource Economics Society.