International Commission of Agricultural Engineering Commission Internationale du Génie RUI'al
13
thINTERNATIONAL CONGRESS ON AGRICULTURAL ENGINEERING
Rabat (Morocco) 2-6 February 1998
13
èmeCONGRES INTERNATIONAL DU GENIE RURAL
Rabat (Maroc) 2-6 Février 1998
Proceedings / Actes
VOLUME 3
AGRICUL TURAL MECHANIZA TION
MECANISATION AGRICOLE
CIGR accepts noresponsibility for the statements made. opinions exprcssed and the maps included in these proceedings.
La CIGR se dégage de toute responsabilité pour les déclarations faitcs, les opinions fonnulées et les cartes reproduitcs dans ces actes.
Edited by:
Professor EI Houssine BART ALI. Agricultural Engineering Department of the Hassan II Institute of Agronomy and \ eterinary Medicine of Rabat - Morocco Mohammed DAOUDL Agricultural Engineer. Consultant, Rabat - Marocco
Published by ANAFID 2, Rue Haroun Errachid. Rabat - Marocco
For the complete set ofsixvolumes ISBN <)981-18~n-()-()
Forvolumc l: ISBN 9<)81-1887-1-9
For volume 2: ISBN 9981-1887-2-7
Far volume 3: ISBN <)<)81-1887-3-5
Far volume-l: ISBN <)981-1887-4-3
For volume 5:ISBN <)981-1887-5-1
Forvolume 6: ISBN <)<)81-1887-O-X
CONTENTS / SOMMAlRE
SECTION 3 : AGRlCULTURAL MECHANIZA TION MECANISA TION AGRICOLE
3.1. National trategies For Agricultural Mechanization Stratégi _-ationafes de Mécanisation
Agriculrural mechanization strategy formulation :Concepts and rnethodologv and the rol of the private sector and the government
L J Clarke (FA.O)
lmernational standards foragricultural equipment 17
H.Ru eli, Ha/m. &Barrie L.Smith (USA)
Evaluation methods toassess thebenefits of precision agriculture techniques in the 39 Italian situation
.\1.Lazzari, F Mazzetto &M. Vaccaroni (Italv)
111eresearch and inquiry of mechanized technology for dry agriculture 51 Z. Shukuan, Z.Jingzhe, L.Jushuang & L. Shenggong (China)
3.2. Mechanization Techniques Adapted toMediterranean and Tropical Rcgions 57 Tcchniqucs de Mécanisation Adaptées aux Régions Méditerranécnnes et Tropicales
Studies on useof improved single and double animal harnesses and Yokes 59 M. P.Singh &Jayant Singh (India)
Letravail du sol dans les oliveraies en milieu aride 67
B. Ben Rouina, M. Yousfi, _M. Mlaouah &A. Omri (Tunisia)
Design and trial of a chisel plough fitted with parabolic shanks to work as rigid or 77 flexible shanks
F Plegrin , A.Madueno.M. Lopez, D.L. l/a/era, 1.Aguera, .J.R.Jiménez & GL Bianco (.pain)
Tillag and erosion 85
E Cerruto ,S.Failla & G. Schillaci (Italv)
Soil compa tion management byvari-width subsoiler 93
1.1.JOr/& .Salamon (Hungary)
Axle loadand ground pressure responsibility 011thecompaction of tilled soils. Part I : 101 Rootbed
D..Iorajurta. .F Botta &
t.s:
Draghi (Argentina)Axle load and ground pressurc rcsponsibility 011thc ompaction or tilled soils. Pan Il : IO]
Secdbed
c.
F Bouo. D.Jorajuria, & LII Droghi f.·lrgenllno,Comparison orzero. minimum and convcutional ullag ~ tCI11S011whcat vicld in paddy-wheat rotauon
TV .)/ngl1. & BSingh (Indio)
11~
Seedbed characteristics effects on whcat emcrgcncc inconvemional aud conscrvation 121 tillage systems
.·r
flemma! .·1 (Iran)Des outils intelligents pour les lits de semenccs I:n
R.Rouveure &.f-F Billot (France)
Contròle de l'uniformité de I'enrouleur l-n
.'11..·lzlIuggogh (.Iforocc:o)
Design or adccision-support systcm for thc icchnical-cconomical analysis or I.j.l) agricult ural contractors
.1/. l'accaroni, F.Mazzetto. D. Pirola &(;.Gastelli (Italv)
Artificial neural network for Ù1C indircct cvaluation or tractor cugine torquc 1St) ,)'Sorcinelli & G. Peri (Italv)
Astudv or sugar bcet root strength I()7
L·1. Gemtos (Greecei
Fabrication locale de matériel agricole au Marce 175
j':.H Boati, A. Bahri (Marocco) &P.Schulze Lanuners (Gerntanv)
Preliminary trial of selected agriculturaJ machinery Ior small farms or Morocco I XI T Takesono, l.Jenane &A. Ezzahouani f.l!oroCCO)
Straw-grain separation with screw conveyor in conveniional combincs I X7 A. Ince & E Guezel (Turkev)
Ccrcal harvesting in developing countries: Evaluation or different t hnological 105 solutions
.'1/Cicoria, Ci. Ottaviani, M. Vieri & .11. /.oli (ltalv)
Design. Dcvelopment and testing adate harvcsting machine 205
.1I.Shanisi, .J.Kilgour, R.J. Godwin & S Blackutore (i'm/ed .ingdom)
Development of a new combinc harverster for multi-crop in Japan 2II T Sugtyama, T Ichikawa. Y.Hidaka.,,1. Yamamoto, A.'. Hamada, T Odaliara &
M, Mizumoto (Japan)
Perforrnances d'une arracheuse de pommes de terre 2 19
EH Baali (Morocco) &HJ. Heege (Gennanvt
3.3. Equiprnent and Technical Itincrarics Rcspcctuous of thc Envirunmcnt ll5 Matériel ct Itinéraires Techniqucs Rcspectueux de l'Environncment
Les effets de trois itinéraires techniques sur le comportement du tracteur/outil et sur 227
létat structural du sol
T.Xlansouri, .ILI. BenAbdellah & .II.E Hamza (Tunisie)
Evolution or dry bulk density inavertisol after passes with tracted and wheeled vchicles 237 f)'!', l'o/era,.I Gil, .II.Galvez, .'1.Xladueno. .''L Pena &.I.,t Salinas (Spain)
Exploitation or Diesel internaI combustion engines inihebuildings with lirnited 2.J.5 airexchange
.LX Kartashevich. 1>1.Belousov &.Lt Sushnev (Belarus)
Design and testing of azero emission tractor 253
L Bodria, R. Guidetti & C. Bisaglia (Italv)
Work quality or a fcrtilizer spreader equipped with aDPAE control system 2(l'\
.11.Fio/a & R. Oberti (lta/vi
Development of a robotic sprinkler hcad [or precision irrigation 275
C. Turker (Turkey), B.S:Blackmore &F.
t;
1I'eatherhead (Unhed Kingclotu)Spray volume rate errors in intcnnittent opcration of hydraulic nozzles 283
st.
Sa/vani ([j.~>lJImplementation of a spray distribution modcl for the evaluation or field sprayer boom 29I behavior
Ll,ardoux, CSinfort, A. Miralles, H Bonicelli &F Sévila (Francei
Autornatisation d'un banc de répartition destiné éÌ tcster les buses depulvérisation 305
E. Hamza (Tunisia), F Lebeau & .1 l-I"Destain (Belgium)
praycrs used ingreenhouses: Parameters or distribution J13
E Cerruto, R. D'Antico, S Failla & G Maneuo (lta/vi
'ind tunnel simulation of chemical antidrift adjuvants for pesticide rrcauncnts 325
C. DeZanche, D. Friso, C Baldoin ce- .1.Zelante (ltalv)
or
conservation tillage on watershed hvdrology insemi arid Kcnya: An 335tion of AGNPS. SCS-CN and rational formula runoff models
_ - Btantah, (Kenva), L Stroosnijder {i\'elher/andsi, Te. Sharnia &R.K.J.:.Cherogonv
'61.'0)
III
Work Quality or'a: Fertilizer Spreader Equipped with a DP AE Control System
Marco Fiala and Roberto Oberti •
ABSTRACT
The field performance of a spreader depends on both transversal and longitudinal uniformity of the fertilizers distribution. On the basis of the experimental tests on a DP AE control system (distribution proportional to work speed), carried out with 3different fertilizers, at 3application rates and at 8 speeds (6to 13 kmIh), a model to calculate the total Coefficient ofVariation ofthe spreader was defined, determining a global index of its work quality. Moreover, knowing the fertilizer responses to the different arabi e crops and quantifying the tota! distribution error, it has been obtained a preliminary evaluation of the economie benefit related to the use of a spreader equipped bya DPAE system.
Keywords: fertilization, rate control, distribution uniformity
INTRODUCTION
Within the more advanced agricultural systems, Precision Farming (PF) is the topic on which much of the current scientific research in Agricultural Engineering has been focused (Stafford ed., 1997). This is due to the development of technologies able of localising machines in tbe field, controlling their operations and processing information, thereby enabling site-specific controloffield operations.
However, the widespread adoption of PF is subordinated to several technical and economie factors; this points out the insufficient maturity ofPF and hinders itsrapid commerci al diffusion.
There are also particular situations - such as those existing in lta!ian agriculture - whose structural characteristics (fragmentation ofplots, shape and layout of fields, disformity of crops, farm organisation, etc.) consti tute further causes against the adoption of this new system of"agricultural management" (Mazzetto etal., 1997).
Consequently, itisstili ageneral interest to analyse the performance offarm equipments used in the traditional agriculture, for which the smallest unit of reference is the whole plot. In this connection, one of the most interesting aspects is the evaluation of the work quality improvements which can be obtained through the adoption of innovations (sensors, control systems, etc.), not necessarily integrated intoaPF system.
This paper analyses the main causes of errors in the application rate of fertilizers, and the impact of such errors on the economie performance of a particular crop. However, this approach disregards thesystematical andlor instrumental errors connected to the measurement of the variables.
FERTILIZER CROP RESPONSE, FERTILIZER RATE AND CROP PROFIT ABILITY
For certain types of crops (cereals, roots and tubers), several studies (KIing, 1986; England, 1986; Shiles, 1989; Morrison et al., 1980; Sequi ed., 1997) correlate -for nitrogen only - the crop yield Y (t/ha)with the fertilizer rateD(kg/ha) (Fig. 1).
Authors are Agronomist and Physicist, respectively. University ofMilan, Institute of Agricultural Engineering, ltaly
12.0 'O.
i
l
-,I
corn1
9.0
t
f · , · ·..··; · ·,j
I
ro
1
~ 6.0
t
corn2
--~---+---~
300
100 200
Applicatian rate(N,kg/ha)
Fig. l - Yield response vs.nitrogen fertilization
The various Y(O) functions, notwithstanding common characteristics (concavity towards the bottom, a single maxirnum YM at a rate OM,a value Y(O)>O),exhibit marked numerical variability even within the sarne crop; this clearly limits the applicability of these functions to the site and conditions ofthe experimental measurements.
Nevertheless it is possible to make a few general considerations. In fact, isolating the contribution of fertilizer rate from the other production factors, the protit per unit of surface (specitic protit Ps;ECU/ha) obtainable for a specific crop is given by:
Ps(O)=Y(O)·pu -O-cu [ECU/ha] (I)
where Pu(ECU/t) is the sale price ofthe product andCu (ECU/t) is the cost ofeach unit offertilizer.
The optimal fertilizer rate Do (kglha) must verify the condition dP~ = O; differentiating
(Eq.l) the opti mal fertilizer rate is obtained when:
, Cu
Y(O) =-- Pu
(2)
Do istherefore the fertilizer rate for.which the derivative of the yield response function is equal tothe ratio between fertilizer cost and product price (Fig. 2).
264
Y'(D}<
C.
P,~I--__
Ps (D)
i
~
Y'(D}
c.
P,+---+--="""-'~~
P,;(D)
D L----:!Q,:--->D
Fig. 2 - Qualitative yield response functions to nitrogen fertilization for cereals (A) and roots/tubers (B); identification of the optimal fertilizer rate Do and the maximurn yield fertilizer rate DM;
effect ofthe applied fertilizer rate on the specific profit P,obtainable from the crops.
CAUSES OF ERROR IN THE FERTILIZER RATE
Assurning that the fertilization requirements of a plot are homogeneous -i.e.that the fertilizer application rate DI(kg/ha) must be the same over the whole plot area A (ha) - by operating with a centrifugai spreader with applicationldistribution components properly functioning (constant fertilizer flowrate Q,kg/min; constant spinner rotation speed co,rmp), the possible causes which can lead to an error in the fertilizer application rate DIare:
i. irregular spreader pattern;
ii.incorrect forward travel-line (trajectory) along the field;
iii.variations in the work speed.
For evaluating the cost/benefit ratio of a technological innovation aiming to minimise such errors isessential to quantify their effect on the economie performance of'the crop.
NON-UNIFORM FERTlLIZER DISTRIBUTION AND CROP PROFITABILITY
Subdividing the plot surface A (ha) into elementary areas - whose dimensions (éx; òy) are chosen so that two adjoining areas exhibit anon-uniform fertilizer distribution - the specific profit of the plot A-defined by (Eq.l )-can be expressed as the mean ofthe individuai specific profits of each elementary area:
[ECU/ha] (4)
In(Eq.4), the distribution ofthe fertilizer rate arount its mean value D has no effect on the terrn (D. cu), although -being Y(D) *'Y(D) -it does influence the terrn (Y(D). Pu).
Iff(Dj) isthe statistical distribution of the fertilizer rates in the individuaI elementary areas, (Eq.4) becomes:
"cl
Ps(D)=Y(D)·pu -D.cu =Pu ·L)(DJY(Dj}-D.cu
i=l
[ECU/ha] (5)
where: Ilclis the number ofc1assesintowhich the fertilizer rate range is subdivided.
Equation 5expresses the concept, observed also in practice, that the specific crop profit Psis influenced by the yield and, therefore, by the uniforrnity with which the operator distributes the chosen fertilizer rate.
It istherefore necessary to analyse the behaviour of f(Dj) in relation to the causes of error previously identified.
Irregular Fertilizer Spreader Pattern
The application of a specific fertilizer with a specific spreader has an associated "unitary distribution pattem".
The shape and uniforrnity ofthis pattem depend on the adjustments (ground height, speed of rotation, fin length and angle, etc.) made by the operator on the spinners in accordance with the Manufacturer's recommendations.
The pattem isdeterrnined experimentally bymeans of standardised tests (ASAE S341.2; ISO 5690/1) conducted atvarious testing centers (CEMAGREF, NLH, Statens Maskin-Provningar etc.).
The overall uniforrnity ofthe lateral distribution - obtained from the partial overlap of several adjoining passes ("multiple profile") - is expressed bya Coefficient ofVariation, given by:
(Jq
CV =~·100
p q (6)
where: (Jq(kg/min) is the standard deviation ofthe fertilizer flow rate ofthe collected samples and q (kg/min) is the mean ofthese values.
The reference values commonly used for CVp(Balsari et al., 1994; Yule et al., 1996) are shown in Table 1.
Table 1 - Lateral distribution Coefficient ofVariation
UNIFORMITY of COEFFICIENT of
LATERAL VARlATION
DISTRlBUTION
(%)
Excellent O~CVo~ 5
Good 5<evo
s
lOFair 10<CVo~ 15
Inadequate 15<CVn ~ 20
.Poor CVo> 20
For the purposes ofthis study, the unitary distribution pattem -deterrnined by the operator's adjustments and regulations - is assumed to be constant for ali the passes which make up the complete field fertilization operation.
266
lncorrect Forward Travel-line (Trajectory) Along the Plot
There exists a useful distance bu(m) between adjoining passes which overlaps two unitary distribution patterns in such a way as tominimise the Coefficient ofVariation CVp• This distance b, isdetermined onthe basis ofthe distribution pattern, usually through iterative numerical calculations (Balsari et al., 1994;NLH, 1996).
Inthis way anoptimal travel-line, usually rectilinear, isdeterrnined; the spreader must follow itin order to obtain ali the operation passes equidistantly spaced bybu.
In practice, however, theoperator isunlikely to maintain thé optimal travel-line for the entire fertilization operation; therefore, this study assumes that the actual field position of the fertilizer spreader exhibits anormal distribution around theoptimal distance b.,
The probability that the spreader islocated, at any given time, at adistance b(m) from the optimal trajectory isgiven by:
b2
f(b)= ,I .e-2"~
,,27t ·crb
(7)
The travel-line error b has a negative effect on the uniforrnity of distribution, generating a non-optimal multiple distribution profile.
The magnitude of thevariance (o,2) depends on the possibility ofidentifying and maintaining theoptimal travel-line in theplot (driving ability, field conditions, etc.).
Variations in the Work Speed of the FertiIizer Spreader
Many factors can cause variations in the work speed of a spreader, such as: soil slope and irregularities, shape and size of the plot, slip ofthetractor driving wheels, reversing the direction of travel and the associated manoeuvres, ability ofthe operator.
Because the experimental "data available only quantify work speed variations for sirnilar operations (Grisso et al., 1988), we assume that the speed of the spreader follows a nonnal distribution:
(8)
The value ofthe variance (cr/) depends, obviously, on the magnitude ofthe causes described above under the various operating conditions.
Total Distribution Error
During a fertilizer-spreading operation, the causes of error separately analysed above can coexist, contributing in varying proportion tothenon-uniforrn distribution ofthe fertilizer.
The actual distribution over the entire plot is in fact determined by summing up the longitudinal variability effects (y axis, travel direction) - caused by variations in the work speed of the machine - and the lateral variability effects (axis x), caused by irregularities in the spreader pattem and its incorrect overlap due to travel-line deviations.
The plot A can be subdivided into A, elementary areas of width Sx.(m) and length tJ.y(m).
Inside each A,=tJ.X"tJ.y,values for both the distribution pattem and the work speed are assumed tobe uniforrn.
In practice, the transverse dimension /1x coincides with the resolution used to subdivide the unitary distribution pattem (éx =0.5 - 1.0m), whereas the dimension along the travel direction tJ.y, which takes intoaccount the speeds variations, falls intherange tJ.y=1.0 - 2.0 m.
Inthe case of arectangular field, the total number of elementary areas nA will be:
n =(bLA) .(bLU)
A t.x t.y (9)
where bLA'(m) and bLU(m) are, respectively, the total plot width and length.
By sìmulating the movementof a fertilizer spreader on the field, it ispossible toevaluate the - fertilizer rate Dk (kg/ha) distributed oneach elernentary area Ak.
To thisend, along asingle pass, for each forward step t.yofthe unitary distribution pattem:
Ltheforward speed varies, using random values calculated by(Eq.8).
2. the spreader pattem laterally slùfts from the optimal travel-line, using random values calculated by (Eq.7).
The procedure isrepeated foreach pass of the fertilizer spreader, until the whole plot width bLA(m) has been covered.
DETERMINING THE FUNCTION f(Di)
The fertilizer rates Dk (kg/ha) ineach ofthe elementary areas Ak(ha)aregiven by:
"d
Lmk,j l"d
jel "qk,j
Dk =---=-'L..-- t.x. t.y t.x jel Vk,j
[kg/ha] (lO)
where: na is the number of times (usually 1 or 2) that the unitary spreader pattem "passes over" that Ak; vkj (km/h), mkj (kg)ànd qkj(kg/min) are, respectively, the different work speeds, the rnasses and the fertilizer flow rates in theIlddistributions which pass over elementary area Ak.
Because the travel-line errors of the fertilizer spreader cause a lateral shift in the spreader pattern, it is necessary to redefine - for each trajectory variation .:the fertilizer rates distributed on the underlying elementary areas. For this purpose, by plotting the unitary pattem as a broken-line diagram and 'averaging the values of the broken-line portion that fall upon ~, it is possible to determine the flow rate qkj(Fig. 3).
o ..,
I Al IA2 I A3 I A. I
b
o .1\7
Al I A2 I A3 I A. I
Fig. 3 - Effect of unitary profile slùft (b) on the amount of fertilizer distributed on the various elementary areas AI,....,~, ....
The used algorithm can be summed up in the following steps:
268
(bLA) l. for eachpass from l to
h:- ;
(bLU) 2. foreachforward step from lto 60y ;
3. calculate v(km/h) work speed (Eq.9);
4. ca1culate b(m)deviationfrom optimal travel-line (Eq.8);
5. translate the unitpattern into x=(pass .b,+b); .
6. ca1culate qk (kg/min) fertilizer flow rate for each elementary area Akwhich falls under the unitspreader pattern;
7. ca1culateDk(kg/ha)ftrtilizer rate foreach Akofpoint 6)(Eq.lO);
8. add Dk to the fertilizer rates previously distributed onAk;
9. nextforward step;
lO.nextpass.
This algorithm, implemented using a common programming language, permits the simulation ofvarious operating conditions. These are defined through parameters related tothe spreader (unitary spreader pattern, work speed, nominal fertilizer rate) and to work conditions (speed variations given byO"v;travel-line deviations given byO"b).
The output of the program is the frequency of the fertilizer rate distributed on the nA elementary areas into which the plot was subdivided. This defines the function f(Di), necessary for ca1culating theprofit in(Eq.5).
APPLICATION OF THE CALCULATION MODEL
CentrifugaI Fertilizer Spreader Equipped with an Electronic Controller for Distribution Proportional to Work Speed (DPAE)
Adopting aDPAE controller on a centrifugal spreader should eliminate non-uniformity inthe fertilizer distribution, because the variations in the work speed of the machine are balanced by a proportional variation inthe flow rate of fertilizer tothespinners.
This isequivalent toconsider the effect of speed variations on distribution equals to zero, i.e.
to seto"v=Oin(Eq.9).
However, experirnental tests conducted on afertilizer spreader equipped with a DPAE (Fiala et al., 1997) have shown that the DPAE does not totally eliminate the fertilizer rate errors due to variations inthe work speed. The data coIlected during these tests indicate a Coefficient ofVariation setin the range 2.5%
s
CVvs
6.0%.For the application of th~ simulation model it is possible to assume, for this type of machine, an average work speed v
=
lOkm/h and a CVv=
4.0% (c, = 0.4 km/h).For the other causes oferror, two different operating conditions are simulated (Table 2):
• Case A: excellent IateraI distribution quality (CVp = 5.0%); excellent spreader traveI-Iine O"b
(m
=
0.5 m; CVo.= - =
2.1%);bu
• Case B: poor Iateral distribution quality (CVp = 17.0%); large deviations ofthe spreader from the optimal travel-line (m = 1.5 m; CVo'=6.3%).
A pIot of width bu
=
250 m and length bLU=
250 m, subdivided into llA=
62500 elementary areas (each measuring 60X= 0.5 m by 60y= 2 m), fertilized with nitrogen at the nominaI application rate D, =220 kg/ha by acentrifugal spreader with aworking width of b, = 24 mis considered,Table 2 - Application ofthe model to a spreader equipped with a DPAE in two different operating conditions:Coefficients of Variation
OPERATING CVv CVp
cv,
CONDITIONS
Optimal (CaseA) 4.0% (o-
=
0.4Ian/h) 5.0% 2.1% (CJb=
0.5m) Poor (CaseB) 4.0% (o-=
0.4Ian/h) 17.0% 6.3% (m=
1.5m)The results of the simulations are reported in Figures 4 and 5;they show how uniformity of distribution is affected by incorrect adjustment of the spreader spinners and by deviations from the optimal travel-line. The broader the curve obtained for f(Di), the stronger the negative correlation.
~ 10.0
r
~
~ 8.0",>-
o l
c:
Q)
s.o
!
:::>
CTQ)
~
4.0Q)o c:
~
2.0 :::>oo
O 0.0
o 20 .60 80 100 120 140
Applied rate (% of Dt)
Fig.4-Applied rate occurrence frequency: Case A
40
~
10.0
j
8.0
+
o
>- 1I
oc: 6.0
Q)
:::>
CT
~
4.0Q)o c: 2.0
~
:::>oo O 0.0
o 20 60 80 100 120 1~
Applied rate(% of Dt)
Fig. 5 - AppIied rate occurrence frequency:Case B
40
.cv» =4.0%
cvp= 5.0%
cv ot =
2.1%160 180 200
evv
=
4.0%evp
=
17.0%:CVol
=
6.3%160 180 200
The Coefficient of Variation of the applied rate can provide an index of the totaI work quaIity of a centrifugai spreader:
CVT= aD~ ·100
D (11)
Moreover, the f(Di) obtained from the simuIation permits to calcuIate by (Eq.5) the specific benefit obtainabIe from a particuIar crop.
270
As an example, a field cultivated with com, whose yield response to nitrogen fertilization
Y(D)(t/ha)lies midway between (see Fig. 1):
Yl(D)
= 4.42 + 3.32·1O·2·D- 8.84·1O-s·D2 + 6.60
·1O-8·D3 Y2(D)= 3.94 + 5.56·1O-2·D-3.91·1O-s·D2- 2.22
·1O-7·D3is considered. ,-
Assuming a com sale price pu
=105
ECU/t(excluding tax), a specific cost of nitrogen c,
= 450
ECU/t(excluding tax) and using into (Eq.5) the f(Di) resulting
from the Cases A and Bsimulations, the specific benefits detailed in Table 3 were obtained.
Table 3 - Application of the model to a spreader equipped with a DPAE in two different operating conditions: technical and economie results
OPERATING CVT ECONOMIC YIELD RESPONSE FUNCTION ECONOMIC
CONDITIONS PARAMETERS BENEFIT
(%)
(ECU/t) (t/ha) (ECU/ha)Optimal (Case A) 9.6 p, = 105
Yl(O)=
4.42 + 3.32·1Q"2·0 -760+1160
+ 8.84.10.5.02 +6.60.10.8.03
Poor (Case B) 21.6
Cu= 450
Y2(D)=
3.94 + 5.56·1O·2·D-740+1120
+ 3.9HOS·02 -2.22·1Q"7·D3
Conventional Centrifugai Fertilizer Spreader
For spreaders which are not equipped with a speed controller, it is necessary to evaluate the magnitude of CVv.
In the lack of experimental data, an average work speed of v
=lO
km/hand an operating situation for which the work speed falls in the range 7
~v ~ 13
km/hin 95% of the cases (CVv =
15.0%; crv
=1.5
km/h),were simulated.
In order to compare the results, for both other error causes (Table 4) and operating variables, the same hypotheses of Cases A and B were assumed.
Table 4
-Application of the model to a conventional spreader in two different operating conditions: Coefficients of Variation
OPERATING CVv, CV
pcv,
CONDITIONS
Optimal (Case C) 15.0%
(c-= 1.5
/cm/h)5.0% 2.1%
(CJb= 0.5
m)Poor (Case D) 15.0%
(c-= 1.5
km/h)17.0% 6.3%
(CJb= 1.5
m)The results obtained for Cases C and D are shown in Figures 6 and 7.
10.0
~>. 8.0 oc::
Q)
:::l 6.0 CT
~ s
c:: 4.0~
:::l 2.0 oo O0.0 O
.
,
.v=
evp= 5.0%
cv oì
=
2.1%10.0
~
8.0o
>.
oc:: 6.0
Q)
:::l CT
~
4.0s
c::~
2.0 :::lo o· 0.0 OO
20 60 80 100 120 140
Applied rate (%ofDt)
Fig. 6 - Appliedrate occurrencefrequency:Case C
40 160 180 200
evv
=
15.0%ev p= 17.0%
.Cvol
=
6.3%60 80 100 120 140
Applied rate (%ofDt)
Fig. 7-Applied rate occurrence frequency: Case D
20 40 160 180 200
Considering, for the Cases C and D, the samecrop and the same economie parameters, the specific benefits obtained are shownin Table 5.
Table 5- Application of the model to a conventionaI spreader in two different operating conditions:technical and economie results
OPERATING CVT ECONOMIC YIELD RESPONSE FUNCTION ECONOMIC
CONDITIONS PARAMETERS BENEFIT
(%) (ECU/t) (kg/ha) (ECU/ha)
Optirnal (Case C) 18.1 pu
=
105 Y,(D)=
4.42 + 3.32·1o-2·D + 745.,.1120 - 8.84·1O·s·D2 +6.60 ·10"·0'Poor (Case D) 28.9 Cu
=
450 Y2(D)=3.94 + 5.56·1O·2·D + 725.,.1070 - 3.91·10'·D2 - 2.22·1o-7·D'CONCLUSIONS
A simulation model was developed to determine the non-uniformity of distribution of a centrifugaI fertilizer spreader operating on a generic plot.The non-uniformity was expressed as an applied rate occurrence frequency function f(Di) and as aTotal Coefficient ofVariation (CVT)which provides an total indexofwork quality.
272
The modeltakesinto consideration errors in applied fertilizer rate caused by:a) irregularities inthedistribution pattem; b) incorrect forward travel-line along the plot;c) variations in the work speed.
The algorithm was usedto simulate the behaviour of acentrifugai spreaderequipped witb DPAEcontroller under two different operating situations(Case A:optimal conditions;Case B: poor conditions),obtaining CVTsof9.6% and21.6%,respectively.
Comparison of these indices stresses out the highincidence onwork qualityof botb the distribution pattem (spinners regulation) and the spreader travel-line (passes overlapping) and, consequently, theimportanceto putunderconstant controltbesework parameters.
For each simulated case and on tbe basis of two different yield response functions to nitrogen fertilization,thefunctionf(Di)has been used to calculate tbeeconomie benefit relatedto a com cultivation. Under tbe simulationassumptions, tbe Case B-conditions led, incomparison witb the Case A-conditions, to a reduction of the economic;,benefitfrom 20 to 40ECU/ha.
It is important tostress that, unknowing the funetion f(Di), that means to neglect the non- uniforrnityoffertilizer distribution,theeconomie benefitofthecrop would bewrongly calculated on thebasis oftbe average fertilizerrate(according toEq.I).
The model was alsoappliedto a eonventional fertilizer spreader, assuming eonditions (Case C and Case D) comparable to tbose used for simulating tbe DPAE spreader. The corresponding CVTs, 18.1%(Case C) and28.9% (Case D),were significantly worse than the previous ones.
From the economie standpoint, comparing tbe results witb tbose of the DPAE-equipped spreader,a further reduetion(15+50ECU/ha) inthe erop benefit was aehieved.
The study highlights tbat the work quality of centrifugai spreaders can be substantially improved byadopting systems andlor devices capable of,on the one hand,controlling variationsin work speed(DPA, DPAE) and, on the otber hand, identifying theoptimal travel-line witb greater precision(row-tracing,drivecontrol).
Finally,itshould beemphasised that the experimental determination of the actual workspeed during fieldoperationswould perrnit an even more rigorous calculation of the economie benefits, under variousworkingconditions, connectedwitb adoption ofDPAE systems onfertilizer spreaders.
REFERENCES
Balsari P., Airoldi G., (1994).Spandiconeime(Fertilizer spreaders). Edagricole, pp. 82
Eogland R.A., (1986). Reducing the nitrogen input on arable farms. Journal of Agricultural Economics,37(1):13-24
Fiala M., Oberti R., (1997). Prove di DPAE per il controllo della distribuzione su uno spandiconcime ad azione centrifuga. (Experimental tests of a DPAE control system installed on a centrifugai fertilizer spreader). Proceedings 6th National A.I.I.A. Congress, September 10-12, Ancona,Italy, VoI.3,557-565
Grisso R.D., Hewett E.J., Dickey E.C., Schoieder R.D., Nelson E.W., (1988). Calibration accuracy of pesticide application equipment.Applied Engineering in Agriculture,4(4):310-315 Kling A., (1986). Possibilities and limits of nitrogen fertilization. Economie aspects demonstrated for different crops. Nitrogen,14,FrankfurtJMain
Mazzetto F., Vaccaroni M., Lazzari M., (1997). Agricoltura di precisione: realtà e prospettive.(Precision farming: reality and prospects). Proceeding 6th National A.I.I.A. Congress, Ancona,Italy,September 10-12,VoI.3, 271-280
Morrison J., Jackson M.V., Sparrow P.E.,(1980).The response ofperennial ryegrass to fertilizer nitrogeninrelation to climate and soil.Grassland Researchlnstitute,Hurley,Techhical report n.27 NLH-Iostitutt for Tekniske Fag, (1996).Melding Prove(Nordsten LogicNN),N.3-1996,Norway Sequi P. (ed.), (1997). Progetto finalizzato Produzione Agricola Nella Difesa dell'Ambiente (PANDA).(project "Agricultural productivity protecting the environment"). Agricoltura e Ricerca,
167,5-54; 168,5-100
Shiles R., (1989). Nitrogen for winter barley. Agtec., 1988/89, Hydro Fertilizers Ltd., Ipswich, England
Stafford J. (ed.), (1997). Precision Agriculture '97. Proceedings i"European Conference, VoI. I:
Spatial Variability in Soil and Crop, VoI. II: Technology, 7-10 September 1997, Warwick University Conference Centre, UK, pp. 991
Yule I., Crooks E., 1996. Precision farming: the price of imperfection. Landwards, Spring, 5-9
274
