|
| Front page | | Contents | | Previous | | Next |
Autonomous Weeder for Christmas Trees – Basic Development and Tests
7 Economic feasibility
This content of the previous chapters indicates that it is realistic to apply an autonomous weeder to Christmas trees from a technical point of view. However, it should also be economically feasible. In order
to provide an idea about this the following brief cost calculation was made on estimates for two scenarios: 1) current prices and performances, and 2) prices and performances in 5 years.
7.1 Calculation method
The calculation was performed in two steps. First the machine area capacity was estimated by:
Ah = k v b e
and
As = Ah t n
where:
Ah = area capacity (ha/h)
As = area that can be treated through a complete season (ha/y)
k = conversion factor from m²/s to ha/h
v = vehicle forward speed, m/s,
b = row distance (both sides treated; two cutters), m,
e = field efficiency,
t = average daily working hours, h/d,
n = number of working days available between each weeding
The annual costs of ownership and use of a machine was calculated by:
C = P cf + t n N (m + f + L)
in which:
P = vehicle total purchase price, (DKK)
cf = annual fixed cost factor expressed as a decimal of the purchase price,
N = number of weedings required per year
m = repairs and maintenance, (DKK/h)
f = fuel and oil costs, (DKK/h)
L = labour cost, (DKK/h).
7.2 Chosen parameter values
The most important factors in the calculation are the initial purchase price and the performance of a future commercial autonomous weeder. The usual picture for this kind of technology is that a new product
is introduced at a relatively high price to get the development costs covered as soon as possible. However, after some time, when competitors emerge it is often seen that prices drops, while the technical
performances are improved. To reflect this picture two scenarios were chosen. One is set up with approximate present price and performance levels, while the other is set up with assumed prices and
performances 5 years ahead.
7.2.1 Purchase price
The purchase price (Table 7.1) was estimated on the basis of current prices of the major components and software used for the developed experimental vehicle. No costs was included to cover development
and assembling of the machine. On the other hand the sensors may be bought at much lower prices than estimated when they are to be used in considerable numbers for commercial vehicles rather than for
research.
Table 7.1. Estimated investments needed (2005 and 2010 price-levels in DKK)
| Item |
2005 |
2010 |
| Vehicle platform, incl. engine and basic transmission |
30,000 |
20,000 |
| RTK-GPS incl. Ref. Station |
250,000 |
150,000 |
| Other sensors |
50,000 |
30,000 |
| Computers and electronics |
20,000 |
15,000 |
| Operators PC |
10,000 |
8,000 |
| Software |
20,000 |
7,000 |
| Total Investment |
380,000 |
230,000 |
7.2.2 Weeder performances parameters
The performance parameter values of the 2005 set-up (Table 7.2) are estimated on basis on experience with the ACW assuming that it will be possible to achieve a high operational availability. The values
used in the 2010-scenarie are just rough estimates.
Table 7.2. Estimates of parameter values for capacity and cost calculations
| Variable |
Unit |
2005 |
2010 |
| Fixed cost factor, cf |
1 |
0.14 |
0.14 |
| Number of days between weedings1, n |
d |
10 |
10 |
| Number of weedings per year, N |
1/y |
6 |
6 |
| Vehicle speed, v |
m/s |
0.7 |
1.0 |
| Row distance, b |
m |
1.2 |
1.2 |
| Field efficiency, e |
1 |
0.8 |
0.85 |
| Operation hours per day, t |
h/d |
14 |
18 |
| Repairs and maintenance, m |
DKK/h |
30 |
20 |
| Fuel costs, f |
DKK/h |
10 |
10 |
| Labour costs (20% of normal working hours), L |
DKK/h |
20 |
15 |
1) Operation assumed possible in all weather conditions.
7.3 Results
The results (table 7.3) shows that an ACW built in 2005 would be competitive with the present manual mechanical weeding equipment (current cost price: 3000 DKK/ha), while the anticipated changes in
parameter values during the next five years will appear to bring the costs down to the level of spraying (about 1500 DKK/ha).
Table 7.3. Results of costs calculation
| Variable |
Unit |
2005 |
2010 |
| Hourly area capacity |
ha/h |
0.24 |
0.37 |
| Daily area capacity |
ha/d |
3.39 |
6.61 |
| Seasonal area capacity |
ha/y |
33.9 |
66.1 |
| Annual operation hours |
h/y |
840 |
1080 |
| Fixed costs |
DKK/y |
53,200 |
32,200 |
| Variable costs |
DKK/y |
50,400 |
48,600 |
| Total costs |
DKK/y |
103,600 |
80,800 |
| Area specific costs per year |
DKK/ha |
3,059 |
1,220 |
| Tree specific costs per year (6600 trees/ha) |
DKK/tree |
0.46 |
0.19 |
7.4 Sensitivity analysis
As the selected parameter values only are estimates a sensitivity analysis was made of the following 3 groups of parameter values:
- Machine capacity: v, b, e, t and n
- Capital cost: P and cf
- Variable costs: m, f and L
Table 7.4. Cost sensitivity analysis (2005 scenario)
| Parameter group |
Parameter value Change, % |
Total cost Change, % |
| Capacity parameters |
+ 20 |
- 17 |
| Capital cost parameters |
+ 20 |
+ 10 |
| Running cost rates |
+ 20 |
+ 10 |
| Number of weedings per year |
+ 17 |
+ 9 |
| Number of daily working hours |
+ 20 |
- 10 |
Table 7.5. Cost sensitivity analysis (2010 scenario)
| Parameter group |
Parameter value change, % |
Total cost change, % |
| Capacity parameters |
+ 20 |
- 17 |
| Capital cost parameters |
+ 20 |
+ 8 |
| Running cost rates |
+ 20 |
+ 13 |
| Number of weedings per year |
+ 17 |
+ 10 |
| Number of daily working hours |
+ 20 |
- 7 |
The tables 7.4 and 7.5 show that an increase of the capacity will result in a nearly as large reduction of the costs, while the capacity costs, running cost rates, the number of weedings per year and the number
of daily working hours are less important. It also appear that this picture only is changed moderately with the more favourable parameter values used in the year 2010 scenario.
| Front page | | Contents | | Previous | | Next | | Top |
Version 1.0 November 2005, © Danish Environmental Protection Agency
|