28 issue 2 2007 - crojfe · preventing slip, soil compaction and sinkage, and ... w – dynamic...

2 0 0 728 Issue 2

22

Improvement of wheel skidder tractiveperformance by tire inflation pressure

and tire chains

Stanimir Stoilov

Abstract – Nacrtak

The motion resistance ratio, gross traction ratio and net traction ratio of a wheel cableskidder were determined and mathematical models derived for three tire inflation pressurevalues with or without tire chains on forest road in mountainous conditions.

The motion resistance ratio increases with the increase of tire inflation pressure. On thecontrary, the gross traction ratio increases with the decrease of tire inflation pressure.However, when the tires are equipped with tire chains the skidder motion resistance ishigher. The motion resistance ratio with tire chains also increases with the decrease of tireinflation pressure due to windage between the tires and chains, which results in lowertractive performance of tire chains.

Therefore, by using tire chains, an increase of gross tractive ratio could be achieved, as wellas the increase of motion resistance ratio compared to tires without chains.

The results of the study show that it is strongly recommended to use tire chains as tightly aspossible.

Keywords: wheel skidder, tire inflation pressure, tire chains, tractive performance, motionresistance

1. Introduction – Uvod

The wheel cable skidder is a self-propelled mach-ine designed to transport trees or parts of trees bytrailing or dragging them with one end on the ground.Cable skidders use a main winch cable and cablechokers to assemble and hold the load (Stokes etal. 1989).

Worldwide the wheel skidders are one of thebasic machines in mountain logging on 10–20° steepslopes during thinnings, selection silvicultural sys-tems and shelterwood regeneration.

The increasing demand for more efficient andproductive logging technologies in mountain condi-tions requires improvement of mobility and tractiveperformance of forest machines. One of possible so-lutions is to equip skidders with special high-flo-tation tires with low inflation pressure controlledmanually or by Central Tire Inflation System (CTIS),which allows the increase of tire contact area, thus

preventing slip, soil compaction and sinkage, andimproving traction. Tire chains are often in use tomaximize traction, reduce slipping and fuel con-sumption, and providing minimal brake distancefrom the wheel skidder (Fig. 1). Sometimes the ad-vantages of four-wheel or six-wheel drive skiddertransmission cannot replace the need for tire chainsduring all seasons. In extreme muddy and deep snowconditions tire chains will enhance the mobility andimprove tractive performance, as well as protect thetread and tire side wall from excessive damages.Greater tractive force along with minimum slip pro-vided by tire chains means bigger tree load andhigher speed, and therefore, greater productivity andefficiency of wheel skidder.

The tractive performance, as one of the maincharacteristics of wheel skidders, determines theirmobility, productivity, and environmental damageunder different terrain conditions of forest areas.Tractive performance is in close dependence on soil-

Croatian Journal of Forest Engineering 28(2007)2 137

Original scientific paper – Izvorni znanstveni rad

wheel interaction. The milestones of terramechanicsare the studies of M. G. Bekker (1956, 1969) and J. Y.Wong (1978, 1989). There are numerous papers thatexplored soil-wheel interaction, but investigationsinto use of tire chains have not been given sufficientattention.

The subject of this paper is the research of motionresistance ratio, gross traction ratio and net tractionratio of wheel skidder with or without tire chains.

2. Earlier research – Prija{nja istra`ivanja

The following five dimensionless ratios areused to describe tractive performance (Standards ofISTVS 1977):

Þ travel reduction ratio, commonly called »slip«,is an indication of how the speed of thewheels differs from the forward speed of thevehicle;

Þ tractive efficiency presents the ratio of tractive(drawbar) power divided by correspondingpower at driven wheels (Wong 1998);

Þ gross traction ratio mgt shows the torque de-livered to the wheels divided by the rollingradius of the wheels and divided by the dy-namic load of the skidder:

mgt = T / (r · W) = Fper / W (1)

Þ motion resistance ratio mr describes the re-sistance to movement of a vehicle providedby both the surface on which it moves andinternal friction of its tires, and the energylosses in the wheels divided by the dynamicload of the skidder:

mr = Fr / W (2)

Þ net traction ratio mnt demonstrates tractive for-ce divided by the dynamic load of the skidder:

mnt = Ft / W = mgt – mr (3)

where:W – dynamic load of the skidder

T – torque delivered to the wheels(input torque)

r – rolling radius

The force equilibrium for wheel skidder is asfollows:

Fper = Fr ± Fs ± Fa + Fw + Ft (4)

where:Fper – peripheral force of skidder wheels

Fr – motion resistance force

Fs – slope resistance force

Fa – resistance of inertia

Fw – aerodynamic resistance force

Ft – tractive force (drawbar pullparallel to ground surface)

In mountainous logging the travel speed of load-ed wheel cable skidder is not higher than 6–8 km/h.Therefore aerodynamic and inertial effect could beignored due to low ground speed (i.e. Fa = 0 and Fw = 0).Thus tractive force equation (4) is simplified as fol-lows:

Ft = W · (mgt – mr) = W · mnt (5)

For a four-wheel-drive skidder with rigidly coupl-ed front and rear drive axles, the tractive perfor-mance under a hard surface condition can be achiev-ed if theoretical speed of the front and rear wheels isequal. On forest terrain, where sinkage takes place,the theoretical speed of the front wheels should befaster (i.e. forerunning) than that of the rear becauseof different tire deflection, soil compaction and rutdepth beneath front and rear wheels. To preventtransmission mechanical damage and to reduce tireswear, it is necessary to provide a kinematic discre-pancy between the theoretical speeds of the frontand rear wheels on deformable forest terrain to beused (Dimitrov and Stoilov 2005).

138 Croatian Journal of Forest Engineering 28(2007)2

S. STOILOV Improvement of wheel skidder tractive performance by tire inflation pressure and tire chains (137–144)

Fig. 1 Skidder wheel equipped with tire chains

Slika 1. Skiderski kota~ s postavljenim lancima

There are several studies of tractive performanceof wheel skidders in field conditions (Hassan andSirois 1984, Hassan and Sirois 1985, Sever 1989, Di-mitrov and Stoilov 2005, Stoilov 2005) and labo-ratory conditions (Vechinski et al. 1998, Vechinski etal. 1999). However, only the last two investigationsare focused on the effect of different tire inflationpressures and chains on motion resistance and trac-tion of skidder tires in laboratory conditions.

The research of skidders equipped with tire chainsoperating in logging conditions will be of great im-portance and results will provide better understand-ing of vehicle-terrain interaction as well as ways to in-crease productivity and overall efficiency of wheeledmachines during skidding. All this could be achiev-ed by proper field testing and interpreting the ex-perimental results.

The main objective of this paper is to present theeffects of tire inflation pressure and use of tire chainson the motion resistance and tractive force of a four-

-wheel-drive cable skidder, respectively shown bymotion resistance ratio, gross traction ratio and nettraction ratio.

3. Materials and method – Materijali metode

3.1 Technical data of the studied skidder –Tehni~ki podaci o ispitivanom skideru

The tests were conducted with two four-wheel--drive LKT–81T cable skidders, owned by the PetrohanTraining and Experimental Forestry at University ofForestry. Skidder technical data are shown in Table 1.

3.2 Site and stand conditions – Stani{ni isastojinski uvjeti

The tests were conducted on a flat forest road in a160-year beech stand, located in Petrohan Training


Improvement of wheel skidder tractive performance by tire inflation pressure and tire chains (137–144) S. STOILOV

Table 1 Technical data of LKT–81T articulated four-wheel-drive cable skidder

Tablica 1. Tehni~ki podaci zglobnoga skidera LKT–81T opremljenoga {umskim vitlom formule pogona 4 x 4

Manufacturer – Proizvo|a~ Martimex–Alfa A. S. (Slovak Republic – Slova~ka)

Diesel engine – Dizelski motor 4-stroke, direct injection, turbocharger – 4 cilindra, izravno ubrizgavanje, turbopunja~

Manufacturer – Proizvo|a~ Martimex–Alfa A. S. (Slovak Republic – Slova~ka)

Type – Tip Martin Diesel (Zetor)-8022.138

Rated power – Nazivna snaga 72.25 + 3 % kW @ 2200 min–1

Engine cooling – Hla|enje motora water type – hla|enje vodom

Transmission – Sustav prijenosa snage mechanical – mehani~ki

Gear box – Mjenja~ 5-speed synchromesh – 5 brzina, sinkronizirani

Transfer box – Razdjelnik pogona 2-speed – 2 brzine

Axes – Osovinerigid, mechanical differential lock with electropneumatic control, with final drive

kruta sa zavr{nim reduktorima, mehani~ka blokada diferencijala s elektropneumatskim upravljanjem

Tires – Gume Barum Continental s.r.o.

Type – Tip TL–1 Log Skidder PR 12

Tire size – Veli~ina 16.9–30’’

Tire inflation pressure – Tlak u gumama max. 240 kPa, min. 80 kPa

Skidder mass – Masa skidera 7145 kg

Weight distribution (unloaded skidder)

Raspored mase neoptere}enoga skidera

front axe – prednja osovina (61.5 %)

rear axe – stra`nja osovina (38.5 %)

Winch – Vitlodouble drum, free unwinded cable

dvobubanjsko, slobodno odmotavanje u`eta

Max. line pull – Nazivna vu~na sila 70 kN (bare drum – prazan bubanj)

Max. cable length – Najve}a duljina u`eta 77 m (Ø14 mm)

Drive – Pogon hydrostatic – hidrostatski

Control – Upravljanje hydroelectric – hidroelektri~no

Tire chains – Lanciconventional type with lug rings

uobi~ajeni tip s oslanjaju}im prstenovima

and Experimental Forestry at University of Forestry,situated in Western Balkan Mountain.

Undisturbed soil properties of the site: brown fo-rest soil, average moisture content: w = 20.3%, aver-age cone index: 385 kPa. Atmosphere conditions: airtemperature: 22°C, air moisture content: 61%, andair pressure: 940 hPa.

3.3 Measuring equipment – Mjerna oprema

The test skidder was instrumented with the fol-lowing transducers for measuring tractive force (linepull) Ft (Fig. 2):

Þ a 100 kN load cell for measuring tractive force(line pull) (Dept. of Strength of Materials atTechnical University of Sofia);

Þ a frequency amplifier with 3 ranges: 20, 40and 100 kN (KWS-3073, Hottinger BaldwinMessetechnik GmbH);

Þ a recorder (TSS-101, RFT Robotron).

The measuring equipment devices were power-ed by an autonomous source – two 12 V accumulatorbatteries.

3.4 Study layout – Postavka istra`ivanja

The two-way tests carried out to eliminate pos-sible effects of test area grade (Lta) 30 m in length ofthe road. Tire inflation pressure was set at differentcombinations and the above test sequence was re-peated.

Equal tire inflation pressure was used in frontand rear wheels so as to reduce the effects of kinema-tic discrepancy between the theoretical speeds of thefront and rear wheels on deformable forest terrain.

3.5 Motion resistance measurements – Mjerenjeotpora kotrljanja

The motion resistance force Fr was determined bythe following method (Fig. 3). An auxiliary cableskidder towed the test skidder with a load cell (LC)connected between them. The gear box of the testskidder was placed in neutral position with tire in-flation pressure set at one of the combinations shownin Table 2. At speed lower then 15–20 m/s the motionresistance ratio is assumed to be constant. The mea-sured horizontal force is motion resistance force Fr.



Fig. 2 Measuring equipment

Slika 2. Mjerna oprema

In this case the motion resistance ratio mr is theratio of tractive force Ft = Fr to the skidder weight W:

mr = Fr / W = Ft / W (6)

3.6 Traction measurements – Vu~ni pokusi

The net traction force was determined by thefollowing method (Fig. 4). The test skidder, operatedwith the throttle open wide, towed the auxiliarycable skidder with a load cell (LC) connected bet-ween them. The auxiliary skidder created increasingresistance force by transmission, breaks, and stack-ing blade until test skidder stopped due to 100% slip.The measured horizontal resistance force is maxi-mum tractive force Ft(max).

The gross traction ratio mgt is defined as the ratioof sum of the motion resistance force Fr and maxi-mum tractive force Ft(max) to the skidder weight W:

mgt = (Ft(max) + Fr) / W (7)

4. Research results – Rezultatiistra`ivanja

Motion resistance ratio mr, gross traction ratio mgt

and net traction ratio mnt are calculated from experi-mental data using equations (6), (7) and (3), respec-tively. The average calculation results of the threetire inflation pressure combinations are shown inTable 2 and Figures 5, 6, 7.



Fig. 3 Scheme of motion resistance test

Slika 3. Shema ispitivanja otpora kotrljanja

Fig. 4 Scheme of traction test

Table 2 Calculated results

Table 2. Rezultati prora~una

Tire inflation pressure

Tlak u gumama

Motion resistance ratio

Faktor otpora kotrljanja

Gross traction ratio

Faktor bruto vu~e

Net traction ratio

Faktor neto vu~e

pi, kPa mr mgt mnt

Front

Prednjim

Rear

Stra`njim

without chains

bez lanaca

chains

s lancima

without chains

bez lanaca

chains

s lancima

without chains

bez lanaca

chains

s lancima

230 230 0.0728 0.0798 0.4579 0.5048 0.3851 0.425

210 210 0.0672 0.0800 0.4628 0.5023 0.4021 0.4223

190 190 0.0607 0.0841 0.4695 0.5012 0.4113 0.4171

Fig. 5 Motion resistance ratio vs. tire inflation pressure

Slika 5. Ovisnost faktora otpora kotrljanja o tlaku u gumama

The rut depth after the passes was 50–80 mm. Therut depth difference after front and rear wheels wasnot significant due to macadam bottoming of theforest road.

After calculating these data, statistical processingof mathematical models was carried out. The linear,polynomial, logarithmic and exponential models wereused for describing the relationship between tireinflation pressure and motion resistance ratio, andgross traction ratio and net traction ratio.

The best fitted models describing mr, mgt and mnt

with or without chains are polynomial models shownin Figures 5, 6, 7.

5. Discussion – Diskusija

In case of skidder wheels without tire chains themotion resistance ratio decreases while tire inflationpressure is reduced (Fig. 5). Reduction of tire in-flation pressure causes the decrease in motion re-sistance ratio (7.69%) and increase in gross tractiveratio (2.5%) (Fig. 6) and net tractive ratio (6.8%)(Fig. 7) without tire chains on deformable forest roadat pi = 190 kPa in comparison with pi = 230 kPa. Theseimprovements in tractive performance are on accountof increased wheel contact area and lower sinkagedepth. All these reduce the motion resistance com-ponent (»bulldozing resistance«, Wong 1989) of soildeformation resistance and improve traction forces.

For wheels equipped with tire chains the decrea-se of the tire inflation pressure leads to the increaseof the motion resistance ratio (5.4%) at pi = 190 kPa incomparison with pi = 230 kPa (Fig. 5). The com-bination of tire chains and tire inflation pressurereduction caused the decrease of gross traction ratio(0,7%) (Fig. 6) and net tractive ratio (1.9%) (Fig. 7) atpi = 190 kPa in comparison with pi = 230 kPa.

Reduction of tire inflation pressure leads to thedecrease of the rolling radii of the wheels and wind-age between tires and chains.

Generally, motion resistance ratio of the wheelskidder was higher with chains than without chainsdue to windage between tires and chains caused bythe reduction of tire inflation pressure.

Use of tire chains improves the gross tractive ratiobecause of growth of the traction. Without tire chains,the decrease of tire inflation pressure gives rise to thegross tractive ratio due to the increase of the tire-soilcontact area and adhesion forces, and lower sinkage.

Net traction ratio is the difference between grosstractive ratio and motion resistance ratio. Therefore,on deformable forest road the decrease of tire infla-tion pressure has a favorable impact on the increaseof the net tractive ratio and tractive performance ofthe skidder with wheels without tire chains. Use ofchains and reduction of tire inflation pressure causea slight decrease of net tractive ratio due to highermotion resistance ratio and enlarged windage be-tween the tires and chains.

6. Conclusions – Zaklju~ci

Along with all-drive transmission of wheel skidderand operating factors, the reduction of tire inflation



Fig. 6 Gross traction ratio vs. tire inflation pressure

Slika 6. Ovisnost faktora bruto vu~e o tlaku u gumama

Fig. 7 Net traction ratio vs. tire inflation pressure

Slika 7. Ovisnost faktora neto vu~e o tlaku u gumama

pressure has a noticeable effect on the increase of tirecontact area and improvement of traction, and henceits mobility and productivity.

The results of field tests confirm theoretical find-ings that reduction of tire inflation pressure causesthe decrease of motion resistance ratio and increaseof gross tractive ratio and net tractive ratio withouttire chains on deformable forest road.

The use of tire chains could cause an increase ingross tractive ratio, net tractive ratio, as well as mo-tion resistance ratio in comparison with the casewithout tire chains. Experimental evidence showsthat the use of tire chains leads to the decrease ofgross tractive ratio and net tractive ratio at pi = 190 kPain comparison with pi = 230 kPa, but also to theincrease of motion resistance ratio due to the rise ofwindage between the wheels and tire chains.

The results of the study indicate that the use oftire chains applied as tightly as possible is stronglyrecommended, and that tire inflation pressure mustbe kept in appropriate range during operation toavoid the increase of windage and ensure the opti-mum performance of wheel skidder.

Improved mobility and tractive performance ofwheel skidders equipped with tire chains could de-crease the amount and costs of forest road main-tenance.

7. References – Literatura

Bekker, M. G., 1956: Theory of Land Locomotion. TheUniversity of Michigan Press. 1–499.

Bekker, M. G., 1969: Introduction to Terrain-Vehicle Sys-tems. The University of Michigan Press. 1–520.

Dimitrov, J., Stoilov, S., 2005: Vliyanie na transmisiyatavarhu teglitelno-scepnite svoystva na kolesniya traktor zadarvodobiva (Effect of transmission on wheel skidder trac-

tive performance). Selskostopanska tekhnika (Journal ofAgricultural Engineering) 3: 8–13.

Hassan, A. E., Sirois, D. L., 1984: Rolling Resistance of aCable Skidder – Effect of Inflation Pressure and Tire Sizes.ASAE Meeting Paper No. 84: 10–50.

Hassan, A. E., Sirois, D. L., 1985: Traction and RollingResistance of a Dual-Tired Skidder on Wetland. Trans-actions of ASAE 28(4): 1083–1042.

Sever, S., 1990: Skidder Traction Factors. International Jour-nal of Forest Engineering 1(2): 15–23.

Standards of International Society of Terrain-Vehicle Sys-tems, 1977: Journal of Terramechanics 14(3): 153–182.

Stoilov, S., 2005: Izsledvane na teglitelno-scepnite svoy-stva na kolesniya traktor za darvodobiva (Research ofwheel skidder tractive performance). Dissertation, Uni-versity of Forestry, Sofia, Bulgaria.

Stokes, B. J., et al., 1989: Glossary of terms used in timberharvesting and forest engineering. Gen. Tech. Rep. SO-73.New Orleans, LA, USDA, Forest Service, Southern forestexperimental station.

Vechinski, C. R., Johnson, C. E., Raper, R. L., 1998: Eva-luation of an empirical traction equation for forestry tires.Journal of Terramechanics 35(1): 55–67.

Vechinski, C. R., Johnson, C. E., Raper, R. L., McDonald, T.P., 1999: Forestry Tire Tractive Performance: New, Worn,and with Chains. Appllied Engineering in Agriculture15(4): 263–266.

Wong, J. Y., 1978: Theory of Ground Vehicles. John Wiley,N.Y.

Wong, J. Y., 1989: Terramechanics and Off-Road Vehicles.Elsevier, Amsterdam.

Wong, J. Y., et al., 1998: Optimization of the Tractive Per-formance of Four-Wheel-Drive Tractors: Theoretical Ana-lysis and Experimental Substantiation. Proceedings of theInstitution of Mechanical Engineers, Part D, Journal ofAutomobile Engineering 212(D4): 285–297.

Sa`etak

Pobolj{anje vu~ne zna~ajke skidera promjenom tlaka u gumamai primjenom lanaca na kota~ima

Potreba za ve}om djelotvorno{}u privla~enja drva u planinskim podru~jima zahtijeva pobolj{anje kretnosti ivu~nih zna~ajki {umskih vozila. Kota~ni je skider naj~e{}e {umsko vozilo koje se upotrebljava pri privla~enju drvau planinskim podru~jima. Opremanje skidera {irim gumama te manjim tlakom u gumama omogu}uje pove}anjedodirne povr{ine kota~a i tla, smanjenje klizanja kota~a, zbijanja tla, nastanka kolotraga te pobolj{ava vu~nuzna~ajku vozila. Opremanje kota~a skidera lancima (slika 1) tako|er pove}ava vu~nu zna~ajku skidera uzsmanjenje klizanja kota~a, potro{nje goriva i puta zaustavljanja. U uvjetima slabe nosivosti tla ili visokogasnje`noga pokriva~a uporaba lanaca na kota~ima omogu}it }e ve}u kretnost skidera i vu~nu silu te manje klizanjekota~a, {to }e se o~itovati u ve}im obujmima tovara i brzinama kretanja zbog ~ega }e se pove}ati proizvodnostskidera.



Cilj je rada istraìvanje faktora otpora kotrljanja, bruto i neto vu~e skidera ovisno o uporabi lanaca na gumamakota~a i promjeni tlaka zraka u gumama.

Faktor se otpora kotrljanja opisuje kao otpor pokretanja vozila zbog stanja podloge, unutra{njega trenja gume ienergetskoga gubitka kota~a u odnosu na dinami~ko optere}enje kota~a. Zbog djelovanja zakretnoga momenta nakota~u javlja se obodna sila (Fper) koja sluì za svladavanje otpora kotrljanja vozila (Fr), a ostali dio sile (Ft) za vu~utereta, svladavanje nagiba (Fs), otpora inercije (Fa) i otpora zraka (Fw). Faktor bruto vu~e opisuje se odnosomobodne sile i optere}enja kota~a, a faktor neto vu~e odnosom vu~ne sile i optere}enja kota~a.

Istraìvanje je provedeno na ravnom dijelu {umske ceste, pa nije bila potrebna sila za svladavanje nagiba, azbog malih brzina kretanja optere}enoga skidera izostali su otpor inercije i otpor zraka.

U istraìvanju se koristio skider LKT 81 T ~iji su tehni~ki podaci prikazani u tablici 1. Tlak se zraka u gumamatijekom istraìvanja postavljao na vrijednosti 190 kPa, 210 kPa i 230 kPa. U pojedinom je vu~nom pokusu tlak biojednak u prednjim i stra`njim gumama kota~a.

Otpor se kotrljanja mjerio povla~enjem ispitivanoga skidera u praznom hodu ili neutralnim poloàjemtransmisije te dinamometrom u~vr{}enim izme|u skidera i vozila koje ga vu~e (slika 3). Odnos zabiljeène sile nadinamometru i teìne vozila pokazuje vrijednost faktora kotrljanja.

Vu~na se sila mjerila povla~enjem pomo}noga vozila pomo}u ispitivanoga skidera s u~vr{}enim dinamo-metrom na vu~nom uètu vitla. Pomo}nim se vozilom pove}avala sila otpora ko~enjem, transmisijom i spu{tanjemprednje odrivne daske sve do zaustavljanja ispitivanoga skidera zbog 100 % klizanja kota~a. Odnos zabiljeène silena dinamometru i teìne vozila pokazuje vrijednost faktora bruto vu~e. Faktor neto vu~e za svaki se vu~ni pokusodredio kao razlika izme|u faktora neto vu~e i faktora otpora kotrljanja.

Srednje vrijednosti faktora vu~e pri razli~itim tlakovima u gumama kota~a s lancima ili bez lanaca prikazani suu tablici 2. Ovisnosti pojednih faktora vu~e o tlaku u gumama odre|ene su regresijskom analizom podataka (slike 5,6 i 7).

Manji tlak u gumama bez lanaca na kota~ima uzrokuje smanjenje faktora otpora kotrljanja i pove}anje faktorabruto i neto vu~e zbog pove}anja dodirne povr{ine kota~a i tla i manjega propadanja kota~a u tlo. Kod kota~a spostavljenim lancima smanjenje tlaka u gumama dovodi do pove}anja faktora otpora kotrljanja i smanjenja faktorabruto i neto vu~e zbog slabijega prijanjanja izme|u guma i lanaca te smanjenja polumjera kota~a.

Op}enito skider s postavljenim lancima na kota~ima ima ve}i otpor kotrljanja od skidera bez primjene lanaca.Ve}e vrijednosti faktora bruto i neto vu~e u primjeni lanaca na kota~ima posljedica su boljih vu~nih mogu}nostiskidera zbog smanjenja klizanja kota~a.

Rezultati istraìvanja pokazuju zna~ajan utjecaj smanjenja tlaka u gumama na pove}anje dodirne povr{inekota~a i tla i pobolj{anje vu~ne zna~ajke skidera te time na pove}anje proizvodnosti. Na osnovi rezultatapreporu~uje se postavljanje lanaca na kota~ima {to je mogu}e ~vr{}e uz gumu uz odràvanje tlaka u gumama upotrebnom rasponu kako bi se omogu}ilo bolje prijanjanje lanaca i kota~a te osigurala optimalna vu~na zna~ajkaskidera.

Rad pridonosi boljemu razumijevanju odnosa izme|u zna~ajki vozila i zna~ajki tla te sluì kao smjernica zapove}anje proizvodnosti skidera pri privla~enju drva.

Klju~ne rije~i: kota~ni skider, tlak u gumama, lanci na kota~ima, vu~na zna~ajka, otpor kotrljanja



Author’s address – Autorova adresa:

Asst. Prof. Stanimir Stoilov, PhD.e-mail: [email protected] of ForestryFaculty of ForestryDepartment of Technologies and Mechanization

of Forestry10 Kliment Ohridski Blvd.1756 SofiaBULGARIA

Received (Primljeno): October 19, 2007Accepted (Prihva}eno): December 6, 2007

Evaluation of two harvesting systems forthe supply of wood-chips in Norway spruce

forests affected by bark beetles

Tobias Cremer, Borja Velazquez-Marti


For sanitary reasons, spruce trees affected by bark beetles (Ips typographus L.) should be re-moved out of the stand as soon as possible, to avoid the propagation of the beetles to healthytrees. One possibility, to utilize the accruing crown material in a reasonable way (instead ofburning it) could be, to use it as wood-chips for biomass heating plants. The aim of thisproject was therefore to determine the productivity of two harvesting and processingsystems for wood-chips as a joint-product of round wood in Norway spruce (Picea abies L.)forests affected by bark beetles. Two systems with different sorting criteria were studied:processing of sawlogs, pulpwood and wood-chips (System A) in comparison to the process-ing of only sawlogs and wood-chips (System B). In System A, the energy wood was chippedwith a chipper mounted on a forwarder that was working directly in the stand. In System B,the material to be chipped was previously concentrated along the forest road with aforwarder, and a chipper mounted on a truck was used for chipping.

In System A, 0.18 t of dried chips could be harvested per m3 of round wood, and in System B0.26 t of dried chips per m3 of round wood. The cost of chipping in the stand was 4.74 �/m3 ofchips and the cost of chipping along the forest road after transporting the chipping materialby a forwarder was 5.63 �/m3 of chips. Therewith, a cost-covering supply of wood-chips maybe obtained out of such stands. Concerning the ratio of energy input to energy output it canbe said that the systems required 1.5% and 2% of energy output that was obtained using therespective system.

Keywords: biomass, wood-chips, Picea abies, bark beetle


In the last years, governments of EU membercountries have promoted the use of renewable ener-gy sources. One of the main sources for renewableenergy is the combustion of biomass, which is nearlyneutral in the cycle of CO2. Therefore, many bio-mass-heating plants have been constructed. Actual-ly, most of these plants are supplied with residuesfrom the wood industry (Heller et al. 2004). As thisraw material is limited, new resources have to betapped. For example up to now the biomass pro-duced in agricultural and forestry systems has notbeen fully mobilized and used for energetic pur-poses, due to still unsolved technical problems, highcosts or missing information about the potential andquality of such biomass (FAO 1997, FAO 2003, An-

dersen et al. 2005). Therefore, it is necessary to evaluatethe potentials of biomass and its quality comingfrom forestry and agriculture, and especially to exa-mine the technology available for harvesting andprocessing it.

One possibility to obtain woody biomass for ener-getic purposes is the utilization of trees that have tobe felled and removed due to attacks of bark beetles(Ips typographus L.). These operations show specialcharacteristics. Typically, they are small clear cutswith an area ranging between 0.3 and 1.0 ha. This isdue to the fact that the trees, surrounding the in-fected trees, are often affected by bark-beetles, too,although they do not show yet any visible signs ofattack on the surface. A further spread of the beetleshould be hindered by cutting the neighbouring trees.Contrary to conventional harvesting operations, all


Orginal scientific paper – Izvorni znanstveni rad

the crown material has to be taken out of the stand,to deprive the bark beetles’ breeding material. Typi-cally, the material is then burnt, which means veryhigh labour costs and no income at all. Furthermore,the felling of affected trees is more difficult than thefelling of trees that are not affected, as their crownshave less weight. Therefore the trees do not fall aseasy as trees with green, living crowns. So far onlyvery few studies have been made on the harvest oftrees affected by bark-beetles (example KWF 2004).Although many foresters have to deal with theseproblems in their day to day work, hardly any re-commendations can be found as to how to best dealwith them.

Therefore, the objective of the present work wasto compare different approaches for harvesting and

processing trees in Norway spruce (Picea abies L.)stands affected by bark beetle (Ips typographus L.).Hence, the goal was to compare the profitability ofprocessing sawlogs, pulpwood and wood-chips (Sy-stem A) with the profitability of processing onlysawlogs and wood-chips (including the chipping ofthe traditional pulpwood-assortments (System B).The sorting criteria of both systems can be seen inFigure 1.

Additionally, two chipping systems (chipping di-rectly in the stand and chipping along the forestroad) were analyzed in order to prove techniquesthat could be suitable for certain stands. In bothsystems the following parameters were evaluated:productivity, costs and energy balance of the wholesupply chain and the volume of wood-chips thatcould be obtained with the respective system.

2. Material and Methods – Materijali metode

The main characteristics of the stands to be har-vested can be seen in Table 1. In both systems con-ventional (motor-manual) chainsaw felling and pro-cessing and skidder log extraction have been carried


T. CREMER and B. VELAZQUEZ-MARTI Evaluation of two harvesting systems for the supply of wood-chips ... (145–155)

Fig. 1 Processing systems studied (d – the top-diameter of the stem-parts)

Slika 1. Istra`ivani sustavi izradbe drva (d – promjer na tanjem krajudijela debla)

Table 1 Main characteristics of the stands

Tablica 1. Osnovne zna~ajke sastojina

Stand – Sastojina 1 2

Harvesting system – Sustav pridobivanja drva A B

Species – Vrsta drve}a Picea abies (L.)

Area – Povr{ina, ha 0.46 0.56

Medium DBH – Srednji prsni promjer, cm 49 44

No. of trees per hectare – Broj stabala po ha 124 264

Terrain slope – Nagib terena, % 2 5

Skidding distance – Udaljenost privla~enja, m 350 500

Table 2 Main characteristics of the machines

Tablica 2. Osnovne zna~ajke strojeva

Machines – Strojevi Manufacturer and model – Proizvo|a~ i model

Chainsaw – Motorna pila Husqvarna 394XP – Husqvarna 357XP

Skidder – Skider Mercedes Benz Trac 800

Forwarder (System B) – Forvarder (sustav B) Gremo 950R

Chipper mounted on a forwarder (System A)

Ivera~ postavljen na forvarder (sustav A)

ERJO 7/65 RC (»ERJOFANT«) Power – Snaga: 272 kW

Opening – Ulazni otvor: 40 x 67 cm, Chip reservoir – Obujam spremnika iverja: 10 m3

Chipper mounted on a truck (System B)

Ivera~ postavljen na kamion (sustav B)

Man Truck, Wüestling 600 CV

Power – Snaga: 442 kW; Opening – Ulazni otvor: 70 x 120 cm

out. The team was formed by two workers. One ofthem only cut and processed the trees; the otherworker drove the skidder and occasionally support-ed the felling. In System A, the energy wood waschipped directly in the stand using a mobile chippermounted on a forwarder (Figure 2). In System B, the

energy wood was concentrated with a forwarder inpiles along the forest road and then chipped using achipper mounted on a truck (Figure 3). The charac-teristics of the machines used are shown in Table 2.

To evaluate the productivity, all operations weresupported by time studies. The time of effective work


Evaluation of two harvesting systems for the supply of wood-chips ... (145–155) T. CREMER and B. VELAZQUEZ-MARTI

Fig. 2 Mobile chipper working in the stand

Slika 2. Rad mobilnoga ivera~a u sastojini

Fig. 3 Chipper mounted on a truck, working along the forest road

Slika 3. Ivera~ postavljen na kamion pri radu na {umskoj cesti

and the total working time were recorded for eachworker and machine. All times were defined ac-cording to the REFA-guidelines (1991). At the sametime, the volume was measured of all the logs pro-duced during the operation, as well as the volume ofeach container filled with chips.

In addition to this, the following parameters weredetermined for every container filled with chips:

Þ Moisture content of chips (%) (determined ac-cording to prCEN/TS 14774-2)

Þ Ratio of different size fractions (%) (determin-ed according to CEN/TC 335/EG 4)

Þ Calorific value at different moisture contentsand separated for different size fractions (de-termined with a calorimeter (IKA 2000))

Þ Coefficient of wood-chips potential, calculat-ed by the following equation:

li =V

V

i chips

Where li is the gravimetric coefficient of the po-tential biomass for energetic utilization in a systemof i-characteristics; this coefficient is defined as tonsof dry chips (Vi chips) that can be obtained as a by-product by recovering the residues generated fromthe harvest of 1 m3 of conventional roundwood –sawlogs and/or pulpwood (V).

3. Results and Discussion – Rezultatis diskusijom

3.1 Processed timber products – Izra|eni drvniproizvodi

The products obtained in both systems are shownin Table 3: a remarkably lower volume of sawlogsand wood-chips was harvested in System A. This isdue to the fact, that in System A some trees (es-pecially fir – Abies alba Mill.) could be left in thestand, whereas in System B all trees had to be takenout. As in System B a bigger part of the trees’ bio-mass is used for energetic purposes instead of pro-ducing pulpwood as in System A, the coefficient forthe wood-chips potential in System B is 0.26 andtherewith 44.4% higher in comparison to System B,where the coefficient is 0.18.

3.2 Productivity for felling and processing ofsawlogs and pulpwood – Proizvodnost sje~ei izrade pilanskih trupaca i celuloznoga drva

The distribution of the effective work time andthe productivity of forest workers in both systemsare shown in Figure 4. It can be noticed that forestworkers had a higher productivity in System B, al-though the average diameter of trees in System Awas 5 cm bigger than the average diameter of trees inSystem B, and hence a higher volume of wood as-



Table 3 Processed timber products

Tablica 3. Izra|eni drvni proizvodi


Harvesting system – Sustav pridobivanja drva A B

Number of trees – Broj stabala 57 148

Sawlogs (diameter > 48 cm, length 5 m), m3solid

Pilanski trupci (promjer > 48 cm, duljina 5 m), m3oblovine

24.6 16.7

Sawlogs (diameter > 15 and < 48 cm, length max. 19 m), m3solid

Pilanski trupci (promjer od 15 cm do 48 cm, najve}a duljina 19 m), m3oblovine

70.6 189.4

Pulpwood (diameter > 15 cm, length 5 m), m3solid

Celulozno drvo (promjer > 15 cm, duljina 5 m), m3oblovine

11.8 –

Total volume of sawlogs and pulpwood, m3solid

Ukupni obujam pilanskih trupaca i celuloznoga drva, m3oblovine

107 206.1

Total volume of wood-chips, m3loose

Ukupni obujam drvnoga iverja, m3nasipni

88 243

Total mass of wood-chips (oven dry), kg

Ukupna masa drvnoga iverja (suhe tvari), kg19,530 53,940

Coefficient lj, (tons of wood-chips per m3 of sawlogs and pulpwood)

Koeficijent lj, (tona drvnoga iverja po m3 pilanskih trupaca i celuloznoga drva)0.18 0.26

sortments was obtained. This is mainly due to thefact, that in System B no pulpwood was produced.Therefore, less time was needed for debranching,cross-cutting and measuring the assortments. Con-sequently, the processing ratio in System A is 53% ofeffective work, whereas it is only 45% in System B.

The evaluation of working times and the pro-ductivity of the skidder are depicted in Figure 5. Theeffective work-time distribution of the skidder doesnot differ much between the two systems. In SystemA, less time is needed for driving into the stand anddriving back to the piling site, which can be explain-ed with a shorter skidding distance (System A: 350 m,System B: 500 m). Nevertheless, more time is neededfor unloading the skidder in this system. This is dueto the fact, that, by processing pulpwood, threeinstead of two assortments had to be transported,which means less productivity due to a lowervolume per piece and more time for sorting at thepiling site.

3.3 Productivity of the forwarder (only SystemB) and the chipper – Proizvodnostforvardera (samo u sustavu B) i ivera~a

The forwarder, required in System B for con-centrating the chipping material along the forest road,had a productivity of 23.7 m3

loose/h. This rather highproductivity that was fostered by a comparably lowskidding distance (<100 m) is remarkably highercompared to e.g. a study conducted by the KWF(2004) in comparable stands, where a forwarder reach-ed a productivity of only 17 m3

loose/h.

The following chipper mounted on a truck in thissystem reached a productivity of 69.8 m3

loose/h. Thisis slightly higher in comparison to other studiesdone for example by Asikainen and Pulkkinen (1998)or Basse et al. (2002) for chippers with similar cha-racteristics. Asikainen and Pulkkinen (1998) deter-mined a productivity of 55 m3

loose/h, whereas Basse etal. (2002) calculated a productivity of 40–60 m3

loose/h,depending on the average volume of the trees chip-ped. In studies by Deutschländer-Wolff (2006) orSchuler (2007) comparable chipping systems reach-ed (only) similar productivities, although the chip-ping conditions were more favourable in compa-rison to the present study, due to a much higherpre-concentration of the chipping material.

In System A, the productivity of the mobile chip-per working directly in the stand is 36.4 m3

loose/h.Therewith, the chippers’ productivity is higher incomparison to other studies: Lechner et al. (2007),calculated for a comparable chipper in beech-standsthat are ready for thinning a productivity of only22.5 m3

loose/h and in a study of Thor (1996), a similar



Fig. 4 Effective work-time distribution and productivity of forest workers– Motor-manual felling and processing

Slika 4. Raspodjela efektivnoga vremena rada i proizvodnosti radnikasjeka~a pri strojno-ru~noj sje~i i izradi

Fig. 5 Effective work-time distribution and productivity of the skidder

Slika 5. Raspodjela efektivnoga vremena rada i proizvodnosti skidera

chipper reaches a productivity of approx. 30 m3loose/h

in spruce-beech-stands.When only looking at the productivity of the

chipper, it can be said that the productivity of achipper working at the forest road is nearly 50 %higher in comparison to a chipper working directlyin the stand. The higher productivity resulted fromthe fact that the material was very well concentratedin piles along the forest-road and the chipper there-with could work more continuously. Additionally,the assortment that could have been used for pulp-wood was chipped, too, which obviously increasesproductivity, too. Another reason – of course – is thehigher engine power of the chipper working at theforest road (442 kW in comparison to 272 kW of thechipper working directly in the stand), which alsohighly influences productivity.

In Figure 6, the differences between the two chip-ping systems can clearly be seen: the time for mani-pulating the wood is 34.1% when looking at themobile chipper working directly in the stand andtherewith remarkably higher in comparison to thechipper working at the forest road with only 4.9%.On the other hand, the ratio of the time for thechipping itself is significantly higher when lookingat the chipper working at the forest road (58.6%) incomparison to the mobile chipper working in the

stand (38.8%). The working time for driving in andout the stand, that do not exist in System B, arerelatively low (8.8% and 6.3%), due to a rather shortskidding distance of less than 100 m. In the study byLechner et al. (2007) that was already mentionedabove, slightly higher ratios occur, as a consequenceof a longer skidding distance. As the transport of thechips was well organised when working with themobile chipper in the stand, no waiting times forempty containers need to be noted.

On the other hand the chipper in System B waswaiting about 40% of the working time for new,empty containers (Figure 6), which is a very highquota, and still – according to Wittkopf (2005) – thisis a rather usual proportion in practice. The goal ofchipping operations should always be to minimizestanding-times of the chipper and there is still arather high potential for optimising logistics. If thelogistic system were organised in a better way, and ifwaiting-times could be reduced to 10%, the pro-ductivity of the chipper working at the forest roadwould be close to 100 m3

loose/h. Therewith, it canclearly be seen that an optimization of logistics inchipping operations (continuous transport of chipsand delivery of new, empty containers) is crucial.

3.4 Costs for the supply of wood-chips –Tro{kovi dobave drvnoga iverja

The costs for tree felling are related to differentassortments of round wood obtained (saw logs andpulpwood). Therewith, no costs accrue for the harvestof the chipping material and only the costs for for-warding the material out of the stand and the costsfor chipping have to be considered. In the System A,the calculated cost per working hour of the mobilechipper were 150 �/h (including 15% additionaltimes for delays and rests). In System B, the under-lying costs were 65 �/h for the forwarder and 150 �/hfor the chipper mounted on a truck (also including15% additional times for delays and rests). The totalcosts of both systems are shown in Figure 7.

The costs of both systems are comparable:4.74 �/m3

loose in System Aand 5.63 �/m3loose in System

B (including forwarder and chipper). Therewith, as-suming a revenue for wood-chips of 12 �/m3

loose at fo-rest road, a net revenue between 6.37 and 7.26 �/m3

loose

can be gained when producing wood-chips with thepresented systems. Of course, the costs only forchipping of the material are remarkably higher inSystem A. However, as the costs for forwarding thechipping material to the forest road are obligatory inSystem B, System A is a more favourable solution,when taking into consideration overall costs for thesupply of wood-chips. This is even truer, when look-ing at the assortments produced in both systems. It



Fig. 6 Effective work-time distribution of the chipper

Slika 6. Raspodjela efektivnoga vremena rada ivera~a

can be assumed that the mobile chipper in System Awould have a significantly higher productivity if thepulpwood-assortments were chipped, too, as in Sy-stem B. In reverse, the productivity of the forwarderand chipper in System B would both reach a lowerproductivity, if the pulpwood assortments were chip-ped. As costs per m3

loose differ only by 1 �/m3loose, it

can be assumed that the mobile chipper workingdirectly in the stand is a more favourable system. Onthe other side, if costs per working hour of the chip-pers rise up to 250 �/h, costs per m3 chips approxi-mate and differences become only marginal. Still, itshould be taken into consideration that when usinga mobile chipper working directly in the stand, fewermachines are needed and the organizational efforts,etc. are lower.

These results are significant insofar, as the con-ventional wisdom is refuted that chipping in thestand is more expensive and not profitable (Wittkopfet al. 2003). On the other hand, the results gained byLechner et al. (2007) are confirmed. On a cautionarynote, however, it has to be said that skidding dis-tance is a factor that highly affects the productivityof a mobile chipper working directly in the stand: inthis study, skidding distance was less than 100 m.From a certain distance onwards, a shuttle-forwar-der for transporting the chips to the forest road hasto be used. This again leads to higher costs andfosters therewith a chipping of the traditional pulp-wood assortment together with the remaining crown--material. As shown in Figure 4 and 5, the produc-

tivity of forest workers and skidder is lower whenproducing stem wood and pulpwood instead of pro-ducing only stem wood and avoiding the processingof pulpwood. The same is true for the chipper: itsproductivity is remarkably higher, when chippingcrown material and the traditional pulpwood as-sortments. Therewith, costs for felling, skidding andchipping rise when pulpwood is processed as a se-parate assortment. Consequently, it can be estimat-ed, that with a motor-manual supply of pulpwood, acost recovery for this assortment can scarcely bereached. This conclusion is confirmed through a studymade by Köberle (2007), who also states lower costsfor the motor-manual felling and processing of trees,when only stem wood in combination with wood-chips is produced and no pulpwood is processed.However, it has to be clearly pointed out that thisconclusion is not true for fully mechanized harvest,when a harvester is used, as the additional workingtime for processing pulpwood is only marginal for aharvester! In this situation, the processing of pulp-wood is – assuming current revenues for pulpwood– a more favourable solution (see also Cremer 2007,Lechner 2007).

3.5 Wood energy characteristics – Energetskezna~ajke drva

The chips of Norway spruce (Picea abies L.), pro-duced in this study, had a moisture content of 34.7%in respect to wet weight (Mh), and 56.2% in respectto dry weight (Ms). This rather low moisture contentis not surprising, as the trees were standing dead inthe forest for several months before they were felledand chipped. At this moisture content, chips densitywas 217.14 kg/m3

loose and the obtained calorific valuewas 12.35 MJ/kg. Both values are comparable to thevalues determined by Golser (2004). After oven-dry-ing the chips, the density was 141.8 kg/m3 and thecalorific value increased up to 19.33 MJ/kg.

The ratio of different fractions and its characte-ristics that were obtained after sifting the chips isshown in Table 4. It can be noted that the smallerfractions have lower calorific values. This is sur-prising: often, the smaller fractions contain a highproportion of needles and bark (Suadicani and Gam-borg 1999), which have – due to a high content ofextractives – a significantly higher calorific value incomparison to stem wood (Nurmi 1993), which ismainly found in the coarser fractions. On the otherside, the trees in this study were dead for severalmonths before they were felled and most of theneedles and high portions of the bark were alreadyfallen down. Therewith, the calorific value of thesmallest fraction obviously decreased (Suadicaniand Gamborg 1999). Additionally, a high mineral



Fig. 7 Costs for the supply of wood-chips

Slika 7. Tro{kovi dobave drvnoga iverja

content is often found in the smallest fraction due tocontamination with mainly mineral soil, which againdecreases the calorific value.

According to Table 4, it can be said that it is usefulto sift the chips. Therewith it is possible to eliminatethe smallest fractions of the chips and the energythat can be obtained increases by 6%. Hence, theuniformity and quality of the chips increases and thechips can be sold at a better price.

3.6 Energy balance – Energetska bilanca

As the calorific value of diesel is approx. 47 MJ/kgand its density is 680 kg/m3, the calorific value perlitre of diesel is 31.96 MJ/L. In System A, the cal-culated diesel consumption per effective workinghour of the mobile chipper resulted in approx. 40 L/h.In System B, the average diesel consumption of theforwarder was approx. 10 L/h and diesel con-sumption of the chipper mounted on the truck was68 L/h in average. In Table 5, the energetic balance iscarried out. It can be noted that System B requires anenergetic input that is 34.7% higher in comparison

to System A. This is due to the fact that the chipproduction in System B needs two machines (for-warder and chipper) whereas in System A only onemachine (the chipper) is needed. Still, the energeticinput in both systems is generally low: as it can beseen in Table 5, in System A, 1.5% of the energy thatis obtained has to be put into the system to producewood-chips. In System B, slightly more energy hasto be used to obtain the same energetic output. Still,these results strongly support an application of bothsystems for the production of wood-chips from treesaffected by bark-beetles.


Summarizing the results, it can be said that acost-effective supply of wood-chips out of standsaffected by bark beetles is possible. Consequently, areasonable utilization is given of the crown materialthat was burnt so far. When looking at suitable supplychains, the differences in costs are less than 1 �/m3

loose

between the two chipping systems and therewithsurprisingly small.

Regarding the optimal sorting, it can be said thatavoiding the processing of pulpwood (and produc-ing only stem wood assortments and wood-chips)seems to be a favourable alternative when harvest isdone motor-manually. As a consequence, forestworkers as well as the skidder and chipper reach ahigher productivity and therewith cost per piece forstem wood and for wood-chips decreases.

The calorific value of wood-chips did not differsignificantly from other studies. Sifting of chips canbe useful to eliminate the smallest fractions and there-by to increase the energy output. Additionally, thechips can be sold at a better price. In both systems,the energy output of the chips is by far higher incomparison to the energetic input that has to beinvested to produce the chips.



Table 4 Characteristics of different chip-fractions

Tablica 4. Zna~ajke razli~itih ~estica drvnoga iverja

Fraction diameterVeli~ina ~estice

RatioUdjel

DensityGusto}a

Calorific value*Ogrjevna vrijednost*

mm % g/cm3 MJ/kg

63 3.2 0.1821 19.52

45 32.9 0.2391 19.24

16 31.2 0.1996 18.95

5 21.0 0.1929 18.31

3.15 11.7 0.2270 17.32

* oven dried – suha tvar

Table 5 Energetic balance of the production of wood-chips

Tablica 5. Energetska bilanca pridobivanja drvnoga iverja

System – Sustav A B

Fuel consumed for the supply of wood-chips, L/m3loose

Utro{ak goriva pri dobavi drvnoga iverja, L/m3nasipni

1.13 1.74

Energetic input per m3 of chips at natural moisture content, MJ/m3loose

Uloèna energetska vrijednost po m3 iverja pri prirodnom sadràju vlage, MJ/m3nasipni

36.11 55.61

Energetic output per m3 of chips at natural moisture content, MJ/m3loose

Dobivena energetska vrijednost po m3 iverja pri prirodnom sadràju vlage, MJ/m3nasipni

2681.68 2681.68

Energetic balance (ratio energetic input / output), %

Energetska bilanca (odnos uloène i dobivene energetske vrijednosti), %1.5 2

Acknowledgements – Zahvala

This work has been carried out by the Institute ofForest Utilization and Work Science of the Albert-Ludwigs-University of Freiburg (Germany) and wasfunded by the Deutsche Bundesstiftung Umwelt(DBU). The participation of Dr.Borja Velázquez-Martíin this project was enabled through a postdoctoralgrant from the Ministry of Education and Science ofthe Spanish government.


Andersen, R. S., Towers, W., Smith, P, 2005: Assessing thepotential for biomass energy to contribute to Scotland’srenewable energy needs. Biomass & Bioenergy 29(2): 73–82.

Asikainen, A., Pulkkinen, P., 1998: Comminution of logg-ing residues with Evolution 910R chipper, MOHA chippertruck and Morbark 1200 tub grinder. International Journalof Forest Engineering 9(1): 87–95.

Basse, D., Wassermann, H., Nier, J., 2002: Energieholz-produktion. Teil 3: Waldhackschnitzel (Energy produc-tion. Part 3: Chips out of the forest). Forst und Technik, 10,Munich, Germany, 20–22.

Cremer, T., 2007: Optimizing the wood-chips supply chainusing productivity models. Poster for thee 3rd Forest Engi-neering Conference, October 1–4, 2007 in Mont-Tremblant,Quebec, Canada.

Deutschländer-Wolff, J., 2006: Kosten- und Leistungsana-lyse einer seilkranunterstützten Durchforstungsmaßnahmeam Steilhang zur Energieholzbereitstellung (Analysis ofcosts and productivity of thinnings for the provision ofenergy wood in steep slopes using a cable crane). Master-arbeit am Institut für Forstbenutzung und forstliche Ar-beitswissenschaft der Albert-Ludwigs-Universität Frei-burg, Germany, 1–96.

Heller, M. C., Keoleian, G. A., Mann, M. K., Volk, T. A.,2004: Life cycle energy and environmental benefits ofgenerating electricity from willow biomass. RenewableEnergy 29(7): 1023–1042.

FAO, 1997: The role of wood energy in Europe and OECD,WETT-Wood Energy Today for Tomorrow. Rome: FOPW,Forestry Department, 1–87.

FAO, 2003: WISDOM, Wood Integrated Supply/DemandOverview Mapping, Rome, 1–52.

Golser, M., Nemestothy, K. P., Schnabel, R., 2004: Metho-den zur Übernahme von Energieholz (Methods for takingover energy wood). Forschungsbericht, HolzforschungAustria, Wien, 1–151.

Köberle, M., 2007: Vergleich verschiedener Aufarbeitungs-varianten für die Bereitstellung von Waldhackschnitzelnaus Kronenholz unter Berücksichtigung von Wirtschaft-lichkeit und Hackschnitzelqualität (Comparison of dif-

ferent systems for the supply of wood-chips from crownmaterial with respect to profitability and quality of thewood-chips). Diplomarbeit an der Hochschule für Forst-wirtschaft Rottenburg.

KWF, 2004: Tagungsführer der 14. KWF-Tagung (Guidefort he 14th KWF-Meeting). Groß-Umstadt, Germany, 1–87.

Lechner, H., 2007: Integriertes Konzept zur rationellenRohholzbereitstellung als Beitrag zur Sicherung und Opti-mierung der Versorgung der deutschen Zellstoff- und Pa-pierindustrie (INFOR-Projekt Nr. 71) (Integrated conceptfor an efficient wood supply as a contribution to assureand optimise the raw material supply of the German pulpand paper industry). Projektbericht im Auftrag des Verbandsdeutscher Papierfabriken (VDP) e.V.

Lechner, H., Cremer, T., Becker, G., Willems, S., 2007: DieQual der Wahl: Hacken im Bestand oder an der Wald-straße? (Being spoilt for choice: Chipping in the stand oralong the forest road?) AFZ-Der Wald, 6, Stuttgart, Ger-many, 290–293.

Nurmi, J., 1993: Heating values of the above ground biomassof small-sized trees. Acta Forestalia Fennica, 236, 1–30.

REFA, 1991: Verband für Arbeitsstudien und Betriebsorga-nisation: Anleitung für forstliche Arbeitsstudien (Instruc-tions for time-studies in the forest). Darmstadt, Germany,1–244.

Schuler, U., 2006: Ermittlung und Analyse von Kosten-und Leistungskennwerten eines Fäller-Sammlers bei derPflege von Mittelwäldern zur Bereitstellung von Energie-holz (Analysis of costs and productivity of a feller-buncherin harvesting operations in a coppice with standards sys-tem for the provision of energy wood). Diplomarbeit amInstitut für Forstbenutzung und forstliche Arbeitswissen-schaft der Albert-Ludwigs-Universität Freiburg, 1–170.

Suadicani, K., Gamborg, C., 1999: Fuel quality of whole-tree chips from freshly felled and summer dried Norwayspruce on a poor sandy soil and a rich loamy soil. Biomass& Bioenergy 17, 199–208.

Thor, M., 1996: Chipset 536 C stickvägsgaende flisare –tidstudie ach systemanalys (Chipset 536C stickvägsgaendeflisare – time study and system analysis) Skog Forsk, Up-psala, Sweden, 1–16.

Wittkopf, S., Hömer, U., Feller, S., 2003: Bereitstellungs-verfahren für Waldhackschnitzel: Leistungen, Kosten, Rah-menbedingungen (Supply-chains for chips out of the fo-rest: productivity, costs and general conditions). Berichtaus der Bayerischen Landesanstalt für Wald und Forst-wirtschaft, Munich, Germany, 1–82.

Wittkopf, S., 2005: Bereitstellung von Hackgut zur ther-mischen Verwertung durch Forstbetriebe in Bayern (Supp-ly chains for thermic utilization by forest companies in Ba-varia). Dissertation an der Technischen Universität Mün-chen, Fakultät Wissenschaftszentrum Weihenstephan fürErnährung, Landnutzung und Umwelt., 1–209.



Saètak

Ocjena dvaju sustava pridobivanja drvnoga iverja iz smrekovih {umao{te}enih pojavom potkornjaka

Smrekova je stabla o{te}ena napadom potkornjaka potrebno izvaditi iz sastojine {to je prije mogu}e kako bi sesprije~ilo {irenje {tetnika. Pri tome se na malim povr{inama izvodi ~ista sje~a o{te}enih stabala i susjednih stabalana kojima jo{ nisu vidljivi znakovi o{te}enja. Za razliku od uobi~ajenih postupaka pridobivanja drva potrebno jesve drvo (oblo drvo i granjevinu) iznijeti iz sastojine. Dosada{nje spaljivanje drva imalo je visoke tro{kove rada bezfinancijske isplativosti. Stoga se preporu~uje iskoristiti drvo iz sanitarne sje~e za pridobivanje iverja. Cilj je radausporediti dva razli~ita sustava pridobivanja drvnoga iverja iz smrekovih sastojina o{te}enih pojavom potkornjaka.Sustav A razumijeva izradu pilanskih trupaca, celuloznoga drva i drvnoga iverja iz preostaloga drva. Sustav Brazumijeva izradu jedino pilanskih trupaca iz debla, dok se celulozno drvo pridodaje drvu za iveranje. Kriterij jerazvrstavanja dijelova stabala u navedene drvne proizvode u sustavima A i B prikazan na slici 1. Tako|er suanalizirana dva sustava iveranja (iveranje u sastojini i iveranje na {umskoj cesti) radi odre|ivanja prikladnijemetode u uvjetima sanitarne sje~e smrekovih stabala. Pri tome su promatrani ovi parametri: proizvodnost,tro{kovi, energetska bilanca sustava i obujam drvnoga iverja koje se moè proizvesti primjenom odre|enogasustava pridobivanja.

Osnovne su zna~ajke ispitivanih sastojina prikazane u tablici 1. U oba se sustava sje~a i izrada stabala obavljalamotornom pilom, dok su se pilanski trupci i celulozno drvo privla~ili skiderima. U sustavu A drvo se iveralo usastojini mobilnim ivera~em postavljenim na forvarder (slika 2), dok se u sustavu B drvo za iveranje (uklju~uju}icelulozno drvo) prikupljalo forvarderom i slagalo u sloàjeve pored {umske ceste gdje se iveralo ivera~empostavljenim na kamion (slika 3). Osnovne su zna~ajke kori{tenih strojeva prikazane u tablici 2. Za odre|ivanjeproizvodnosti proveden je studij rada i vremena za svakoga radnika i stroj u sustavu te izmjereni obujmi izra|enihdrvnih sortimenata i obujmi kontejnera napunjenih drvnim iverjem. Iz uzorka je svakoga kontejnera odre|ensadràj vlage iverja, raspodjela iverja po veli~ini ~estica, ogrjevna vrijednost i koeficijent iverja odre|en odnosommase suhe tvari koja se moè pridobiti od drvnoga obujma obloga drva.

U sustavu A zna~ajno je manji obujam izra|enih drvnih sortimenata (pilanskih trupaca i celuloznoga drva) jersu pojedina jelova stabla ostala na povr{ini sje~ine, dok su u sustavu B sva stabla posje~ena na povr{ini sje~ine. Usustavu B ve}i se dio stabala koristi za pridobivanje drvnoga iverja te je zna~ajno ve}i ukupni obujam dobivenogadrvnoga iverja (tablica 3).

Utro{ci su efektivnoga vremena rada i prozvodnosti radnika sjeka~a i skidera prikazani na slikama 4 i 5. Radniksjeka~ u sustavu B ima ve}u proizvodnost, iako je u sustavu A ve}i srednji promjer stabala i ukupni obujamizra|enih drvnih sortimenata. Razlog leì u ~injenici da se u sustavu B iz debla izra|uju jedino pilanski trupci te jepotrebno manje vremena za kresanje grana, trupljenje i preuzimanje. Efektivno se vrijeme rada skidera ne razlikujeizme|u sustava A i B. U sustavu A udaljenost je privla~enja manja, ali je ve}i utro{ak vremena na pomo}nomstovari{tu zbog razvrstavanja drvnih sortimenata iz tovara na pilanske trupce i celulozno drvo.

U sustavu A proizvodnost mobilnoga ivera~a iznosi 36,4 m3nasipni/h. U sustavu B proizvodnost forvardera pri

izvoènju energijskoga drva iznosi 23,7 m3nasipni/h, a ivera~a postavljenoga na kamion 69,8 m3

nasipni/h. Ve}a jeproizvodnost ivera~a na kamionu posljedica ve}e snage pogonskoga motora te skupljanja drva za iveranje usloàjeve uz {umsku cestu, ~ime se omogu}uje neprekidan rada ivera~a (slika 6).

Ukupni su tro{kovi po jedinici izra|enoga drvnoga iverja (slika 7) izra~unati na osnovi proizvodnosti iodre|enoga tro{ka radnoga sata, koji iznosi 150 EUR/h za mobilni ivera~ i ivera~ na kamionu te 65 EUR/h zaforvarder (uz dodatno vrijeme od 15 % za sve strojeve). Jedini~ni je tro{ak iveranja manji u sustavu B, ali je ukupnijedini~ni tro{ak ve}i nego u sustavu A zbog tro{ka forvardera pri izvoènju drva za iveranje.

Pri sje~i i izradi stabala preporu~uje se izrada jedino tehni~ke oblovine iz debla te iveranje celuloznoga drva uzpreostalo drvo, {to }e za posljedicu imati ve}u proizvodnost radnika sjeka~a, skidera i ivera~a. Mobilni ivera~ usustavu A moè posti}i zna~ajno ve}u proizvodnost ako se za pridobivanje drvnoga iverja koristi i celulozno drvo.Prednost sustava A tako|er se ogleda u primjeni manje sredstava rada te time u jednostavnijoj organizaciji rada.Na osnovi se rezultata zaklju~uje da je pridobivanje drvnoga iverja metodom iveranja u sastojini povoljniji sustav.



Sadràj je vlage u iverju vrlo nizak (34,7 % u odnosu na masu svjeè tvari, tj. 56,2 % u odnosu na masu suhetvari) jer je pridobiven iz suhih smrekovih stabala zbog napada potkornjaka. U tablici 4 vidljivo je da manje ~esticedrvnoga iverja imaju manju ogrjevnu vrijednost, {to je posljedica velikoga udjela iglica i kore. Ve}e se ~estice iverjapridobivaju ponajprije iveranjem drva debla te stoga imaju ve}u ogrjevnu vrijednost. Razdvajanje drvnoga iverjapo veli~ini ~estica moè biti opravdano radi pove}anja dobivene energetske vrijednosti i postizanja ve}e cijenedrvnoga iverja.

Za prora~un se energetske bilance (tablica 5) u sustavu A uzela potro{nja goriva ivera~a na forvarderu od 40 L/h,odnosno u sustavu B potro{nja goriva ivera~a na kamionu od 68 L/h i forvardera od 10 L/h. Sustav B tro{i 34,7 % vi{eenergije od sustava A jer zahtijeva uporabu dvaju strojeva – forvardera i ivera~a. No, op}enito je u oba sustava malenutro{ak energije s obzirom na energetsku vrijednost dobivenoga drvnoga iverja.

Klju~ne rije~i: biomasa, drvno iverje, Picea abies, potkornjak



Authors’ addresses – Adresa autorâ:

Tobias Cremer, MSc.e-mail: [email protected] University FreiburgInstitute of Forest Utilization and Work ScienceWerthmannstraße 679085 Freiburg i. Br.GERMANY

Borja Velazquez-Marti, PhD.e-mail: [email protected] University of ValenciaDepartment of Mechanization and Agrarian

TechnologyCamino de Vera 1446022 ValenciaSPAIN

Received (Primljeno): September 28, 2007Accepted (Prihva}eno): November 19, 2007

Influence of working conditionson overlapping of cutting and ground

skidding in group work

Anton Poje, Igor Poto~nik


This research deals with the influence of working conditions on the duration of cuttersparticipation in group work skidding. Two main theoretical assumptions were considered.First, working conditions that decrease the skidding efficiency increase the time of cuttersassistance, because the group as a whole strives to improve efficiency, and second, theduration of assistance to the working group is specific regardless of working conditions. Theresearch included 5 places of research (compartments) in Slovenian state forests, with3 working groups consisting of two cutters and a cable skidder driver. All working groupsused the cable skidder IWAFUJI–T41. The model included 100 cycles of downhill timberskidding. We have established that the duration of cutters assistance is reciprocal to theskidding distance, average bunching distance and number of logs in the load. The modelaccounts for 40.7% of the whole variability. After excluding the influence of workingconditions between the groups, there are no typical differences in the duration of cuttersassistance. Results imply that the number of working group members should be adjusted toworking conditions. On the other hand, insignificant differences of cutters assistanceduration between the groups show that the mere group daily standard is a motivationalfactor big enough for the rationalization of group work of workers who are not additionallyeducated and qualified for group work.

Keywords: group work, cutting, skidding, working conditions, multivariate analysis


Regardless of new technologies, such as cut-to-length logging, the traditional way of logging (use ofchainsaw and cable skidder) will remain an import-ant working technology in future, especially in theforests with difficult terrain conditions, in smallerprivate forests where personal work represents theessential profit for the owner, and in forests withsmall private patches of land. The fundamental cha-racteristic of group work, as one of the organiza-tional forms of traditional logging, is the group workwith at least two members, who simultaneously dothe cutting and ground skidding, and also assisteach other. We estimate that this kind of methodrepresents the prevailing organizational form, ac-cording to the amount of work and quantity of timbercut in Slovenian state forests.

In practice the size of working group dependsespecially on the capacity of skidding means. There-fore, the working groups are usually bigger with theuse of cable skidders as opposed to the use of adjust-ed farm tractors. The reason lies in the relativelyhigh price of skidding means. For the purpose ofincreasing the utilization of skidding means, theworkers in a group have a group daily standard(expected daily efficiency) and receive piece-rate pay-ments. Thus, the workers, as circumstances require,are »forced« to participate in a working phase thatrepresents the bottleneck for the optimal group workefficiency. Therefore, the working time structure andoverlapping of cutting and skidding processes changefrom group to group (Klun and Poje 2000).

With this research we have tried to add some newknowledge to relatively poorly studied relations be-tween cutting and ground skidding in group work,and also highlight the problem of group work form



in altering working conditions in the forest, with thepurpose of including the ergonomic and economicwork organization as additional starting points.

2. Previous research – Dosada{njaistra`ivanja

In Slovenian forestry, the group work has beenintroduced to cutting and ground skidding processin the early 1970s. With complete mechanization oftimber skidding that occurred around 1970, the trac-tor work efficiencies started to decrease, especiallywith cable skidders. Thus, the precondition for form-ing the new working processes was made, suggest-ing cutting and skidding group work. The groupwork demands a highly skilled worker, who is ableto perform any kind of work in the group. Thisrequires additional and constant education, and alsoemployment stability for workers. This can only beachieved with better evaluation of forest work (Kri-vec 1979).

Since 1979, different authors have theoreticallyand practically considered group work from dif-ferent perspectives. Kri` (1984) describes the reasonsfor introducing group work and group work expe-rience in the company Posestvo Sne`nik. Organi-zational forms of group work have changed fromI+2 (1 tractor driver, 2 cutters), I+3, to II+6, and inextreme cases up to II+12. The reasons for intro-ducing group work were the increased timber cutstocks in forests and unused co-worker in skiddingprocess; whereas the reasons for increase of the latergroup were timber volume composition at the road-side landing and additional but unused group leader.Kerne`a (1999) in his later research in the same com-pany, states that the efficiency of cutting and skidd-ing with adapted farm tractors in smaller groups hasincreased with respect to larger groups, whereas thetrend is quite the opposite in the skidding processwith cable skidders.

Cimper{ek (1987) emphasizes that the reliabilityof working means is the most important productionfactor and also adds that the workers’ premium paybased on quantity and disregarding the quality, eco-nomy and damage decrease has an adverse effect onthe worker. The work performance in this case is aroutine one. Only after the change of remunerationmode the process of mass innovation can begin.

Korbar (1988) theoretically discusses the formand conditions of group work organization, and ad-vantages and disadvantages of this kind of work. Asessential advantages he highlights humanization, syn-chronization that enables the implementation of fo-rest management plans and shortens the time fromthe cut to the timber sale. The work gets intellectual-

ly more demanding and enables workers’ creativityand innovation, strengthens the sense of responsibility,increases performance and improves the worker’s re-lationship to forest and working means. The condi-tions for group work are proper working means,quick flow of material and information, physicallyand psychically fit workers with adequate know-ledge.

By comparing individual and group cutting andskidding work, Kruh (1989) states the rationaliza-tion of chainsaw working time, unproductive time,time for wood bunching from the point of view ofthe cable skidder, and also the whole time requiredfor cutting and skidding. He defines the decrease ofunproductive time as the disadvantage of group work,because it decreases the time for breaks and otherworker’s needs.

Winkler (1990) sees the optimal form of workinggroup as the complex working group that performsall works on the permanent, smaller forest surface,with strict cutting and skidding time synchroniza-tion with silviculture performed only in specific timeperiods.

When determining the standard times for theorganizational form I+2 and skidding timber withcable skidder IWAFUJI T–41, Klun and Poje (2000a,2000b) establish that the efficiency of wood bunch-ing and work on roadside landing is lower in com-parison with traditional organizational form I+1 (onedriver and one helper – hooker). The reason lies inthe assistance of tractor driver to cutters and addi-tional time required for timber classifying at theroadside landing. Indirectly, they also establish asfollows: cutters participate in ¾ of all recorded cyclesof the skidding process; according to working opera-tions where the cutters participate in the skiddingprocess, each group has a specific working method;the time cutters spend participating in a bunchingprocess during the timber skidding at the same timedepends on the number of logs in the load, theaverage piece volume and the distance in wood bunch-ing process, and separately on the skidding distance.

Authors state that the need for introducing groupwork has changed through time. While in the 1970sthere were tendencies for the humanization of work,today the aim is to decrease expenses and increaseflexibility, and thus enable the increase of innovationflow in companies and working processes. There-fore, the group work seems ideal for raising pro-ductivity and quality on one hand and for decreas-ing expenses on the other hand. By participation offorest workers in planning, organization and con-trol, possibilities arise for the utilization of unusedinnovation and talent potentials (Lewark et al. 1996).With greater participation of workers in company’s


A. POJE and I. POTO^NIK Influence of working conditions on overlapping of cutting .... (157–167)

activities, partial responsibility shifted to workers, andadditional education in German ThüringenForst, thecompany has managed to improve working en-vironment and joy at work, decrease delays andincrease workers’ self control in highly mechanizedcutting and skidding processes. In four years thepositive effects of group work were seen in greaterharvester (44%) and forwarder (49%) utilization(Findeisen 2002).

Utilization of working means influences the opti-mal size of working groups. Ze~i} and Maren~e (2005)have developed a mathematical model for optimalgroup size, conducting cutting, skidding processesand simultaneous timber manipulation on roadsidelandings. They state that with skidding distance from150 m to 650 m the optimal working group consistsof 6 cutters (5.3), 5–7 tractor drivers (5.1–6.7) and1 cutter – group leader.

3. Theoretical basis and research aims –Teorijske osnove i cilj istra`ivanja

Cutting and skidding group work consists of twoworking processes that are, unlike the individualwork, partly overlapping. Apart from safe and ergo-nomically designed work, one of the main aims ofgroup work is also to maximize the effect. It meansthat the sum of cutting time (tc), skidding time (ts),and the time when these two phases overlap (tc–s)has to be minimal (Fig. 1). All times depend onworking conditions and working technology. How-ever, only a part of skidding time is irrespective oforganizational form of group work (or the groupsize), because it is mostly connected to the capacityof skidding means (unloaded drive, loaded drive,winching).

When working technology, organizational formand skidding means are unchanged, the time of bothprocesses overlapping depends on cutting and skidd-ing working conditions. Disadvantageous cuttingworking conditions (smaller trees, dense stands) in

general decrease the cutters involvement time inskidding process and increase tractor drivers’ in-volvement time in cutting process, and vice versawhen the skidding conditions are disadvantageous(long skidding and bunching distances, low timberconcentrations, steep terrain and skid trails). Utiliza-tion rate of skidding means is thus greater in dis-advantageous skidding conditions, because it enablescutters’ help to tractor driver.

The assumptions mentioned above are true, ifgroups consider only working conditions and if theirspecific work realization or work method has noinfluence on the rationality of work. The existingresearch (Klun and Poje 2000) and the fact that groupswere not additionally educated for group work, showthat the differences could exist between groups re-garding cutter’s involvement in skidding process(and also the tractor driver in cutting process), beingmore a consequence of self-taught working methodthan of rational approach.

According to the enumerated theoretical assump-tions considering group work and data adequacy,we limited our research only to analysis of cutters’involvement in skidding process and set two basichypotheses:

Þ Working conditions factors (skidding distance,bunching distance, skidder short movementsdistance during formation of full load, thenumber of logs in the load and the load volume)increase the duration of cutters’ assistance tothe tractor driver.

Þ The duration of cutters’ assistance is differentaccording to different working groups, even ifthe influence of working conditions is includ-ed in the comparisons.

The positive answer on the first hypothesis wouldenable us to optimally form the working group sizeaccording to effectiveness and difficulty of work.The confirmation of second hypothesis would ad-vise us on the need of workers’ education for groupwork and relevance of income or group daily ef-ficiency as a motivational factor.

4. Research places – Mjesta istra`ivanja

The research was limited to 9 places (compart-ments) in Slovenian state forests situated on the alti-tude from 470 m to 1070 m, where the characteristicsof high Karst prevail (Table 1). In lower parts, com-munities Querco-, Blechno- and Hacquetio- Fagetumprevailed, whereas in higher parts sub-associationof forest communities Abieti-Fagetum prevailed. Theaverage volume of tree selected to cut ranged from0.63 to 2.03 m3/tree, and from 1.71 to 4.10 m3/tree in


Influence of working conditions on overlapping of cutting .... (157–167) A. POJE and I. POTO^NIK

Fig. 1 Cutting and skidding group work schemeSlika 1. Skica sje~e i privla~enja drva u skupnom radu

compartments with sanitation cut (subjects 8 and 9).The compartment surface ranged from 32 ha to66 ha; however the research comprised only a part ofthese surfaces.

In the selected compartments we monitored thework of 6 working groups with organizational formI+2, meaning that the tractor driver and 2 cuttersformed a group. For skidding, all groups used cableskidder IWAFUJI-T41 without remotely controlledwinch. The maximal average age and workingexperience difference within groups were 8.7 and13.5 years, 7 respectively.

In general, the cutting process proceeded in sucha way, that the majority of work was done by thecutters. Only in some specific and more difficultoperations, clearing the lodged tree or activities tomeet the forest regulations (pile up branches) forexample, tractor driver assisted them. Regardingskidding process, the tractor driver managed all

working operations, i.e. empty and fully loaded driveand winching logs, and also the majority of remain-ing work operations, all by himself. The cutters as-sisted mainly in wood bunching process. The workparticularity is the additional cutting and timberassortment done at the roadside landing by the trac-tor driver. The workers had to meet the group producti-vity expressed in daily timber quantity on the roadsidelanding (by the forest road).

5. Methods of research – Metodeistra`ivanja

The time studying was done in 29 days, in sum-mer months of 1996 and 1997. During that period, onthe skidding distance of 1000 m, we recorded 218timber skidding cycles, of which 78 under uphillskidding, 114 downhill skidding, and 26 level skidd-ing categories (Table 2). The work was done by tworesearchers, one recorded the duration of individualworking operations by using the snap back timing,whereas the other measured the volume of loads, thecontrolling time, and evaluated the bunching, standmovement and movement of skidder on roadsidelanding distances. The time study was made with4/1000 of minute accuracy, whereas the error betweenrecording and controlling time did not exceed 5%.

Due to considerable difference between the num-ber of recorded cycles by working groups, especiallyby individual skidding categories (Table 2), our analy-sis, out of all recorded timber skidding cycles, in-cluded 100 downhill cycles recorded in groups 2, 3,and 4 on the subjects 1, 3, 4, 5, and 6. The categoriesof skidding were calculated as weighted averageincline in the direction of skidding with skidding



Table 1 Compartment characteristics

Tablica 1. Zna~ajke istra`ivanih odjela

Place

Mjesto

Compartment – Forest management unit

Odjel – Gospodarska jedinica

Surface area, ha

Povr{ina odjela, ha

Altitude, m

Nadmorska visina, m

Mean cut tree, m3

Srednje sje~no stablo, m3

Group

Skupina

1 61 – Grintovec 32.10 450–510 1.05 4

2 121 – Grintovec 43.30 780–870 2.03 1

3 49 – Mozelj 43.74 470–620 0.63 2

4 74 – Mozelj 40.59 520–600 0.86 3

5 143 – Mozelj 40.86 530–620 0.73 3

6 48 – Stojna 56.28 900–1070 1.02 2

7 90 – Gr~arice 65.81 740–830 1.22 6

8 100 – Gr~arice 33.90 800–900 4.10 6

Fig. 2 Group work with skidder IWAFUJI–T41

Slika 2. Rad skupine sa skiderom IWAFUJI-T41

distance as weight (uphill >5%, –5% � level �+5%,downhill <–5%).

The overlap of cutting and skidding in groupwork was the reason for adding two new operationsto »traditional« skidding operations, naming themassistance to cutters and cutters assistance. Due tothe fact, that during recording the tractor driver wasthe priority, the cutters assistance was noted only incase of his inactivity or where one or both cuttersassisted in timber skidding. In theory, the time va-lues of cutters assistance should change according toworking conditions or working methods within thegroup. These values served us as a dependant va-riable in our further analysis.

For the independent variables we used the fac-tors influencing the cutters participation in timberskidding process i.e. the factors from the first hypo-thesis. The skidding distance was measured as thedistance between the spot where the full load wasmade and the spot for timber unhooking (roadside

landing). The average bunching distance was calcu-lated as the average of distances required for bunch-ing of individual logs in the load. Short movementsdistance equals the sum of all skidder movements inthe stand during the formation of full load and de-pends on the timber concentration. The number oflogs in the load and the load volume are calculatedas the sums of individual logs.

6. Results – Rezultati

The comparison of average duration of cuttersassistance among groups (Table 3) showed that thereare differences between the groups (p <0.000). TheWelch test was used for testing the equality of meansbecause the Levene test of homogeneity of variancesshowed that the differences among the variances aresignificant (p = 0.021). By applying the posteriordifference tests among the means (Tukey HSD), wefound out that the cutters assistance differences areonly characteristic between the groups 2 and 3, and 2and 4, whereas there are no statistically confirmeddifferences between groups 3 and 4.

With average values of working condition factorswe can infer that the working conditions differ byworking groups and that they are generally lowerfor the group 3, which has also lower cuttersassistance value in comparison to other two groups.It is necessary to stress here that the average valuescan also indicate the danger of stratified data, notonly by working group but also by the individualfactor. In such case we cannot establish preciselywhether the differences of cutters assistance are theconsequence of studied factor or the specific groupwork method. In our case this happened with theskidding distance, which is the consequence of thesampling method (Fig. 3). Primarily, we used thedata for the standard time calculation and had the



Table 2 Structure of recorded cycles regarding skidding direction

Tablica 2. Struktura snimljenih turnusa s obzirom na smjer privla~enja

Group

Skupina

Number of cycles – Broj turnusa

All

Svih

Uphill

Uzbrdo

Level

Bez nagiba

Downhill

Nizbrdo

1 25 4 12 9

2 33 – – 33

3 56 17 5 34

4 42 – 9 33

5 10 10 – –

6 52 47 – 5

Sum – Ukupno 218 78 26 114

Table 3 Average values and correlation between cutters assistance and working condition factors

Tablica 3. Prosje~ne vrijednosti i korelacija izme|u pomo}i sjeka~a te ~imbenika radnih uvjeta

Group

Skupina

Cutters assistance

Pomo} sjeka~a

Skidding distance

Udaljenost privla~enja

Bunching distance

Udaljenost skupljanja

Skidder moving

Premje{tanje skidera

No. of logs in the load

Broj trupaca u tovaru

Load volume

Obujam tovara

min m m m pcs. – kom. m3

2 9.31 694 11.5 12.3 9.0 3.64

3 4.47 189 8.6 15.8 7.6 2.57

4 6.18 313 14.4 31.4 7.8 3.83

Pearson correlation – rPearsonova korelacija – r(Cutters assistance – Pomo} sjeka~a)

0.332 (**) 0.222 (*) 0.086 0.564(**) 0.096

** Correlation is significant at the 0,01 level (2-tailed) – Korelacija je statisti~ki zna~ajna uz razinu zna~ajnosti 0,01 (dvosmjeran test)

* Correlation is significant at the 0,05 level (2-tailed) – Korelacija je statisti~ki zna~ajna uz razinu zna~ajnosti 0,05 (dvosmjeran test)

only goal to record some cycles every 100 m of skid-ding distance.

Figure 3 shows that the data on cutters assistanceby groups is within different skidding intervals, mean-ing that we can conclude that the cutters assistanceincreases with the increase of skidding distance. Tak-ing into consideration the theoretical basis and re-search projects of other authors (Ze~i} and Maren~e2005), we can state with strong likelihood that theshown cutters assistance differences depend also onthe skidding distance and not just on the workingmethod or organization of individual groups.

Studying the linear dependence among indivi-dual factors and cutters assistance (Table 3), we foundout that the cutters assistance increases with theincrease of all working condition factors, according

to positive value of correlation coefficient (r). Cor-relations are statistically confirmed with skiddingand bunching distance, and with number of logs inthe load. The values of correlation coefficients arerelatively low, thus indicating the weak correlationof cutters assistance with working condition factors.Correlation values among working condition factorswere 0.306 at the most, enabling the use of all factorsin multivariate regression. By applying linear multi-variate regression, we tried to establish the influenceof individual working condition factor on cuttersassistance with the constant value of other factors.

By applying the stepwise method of multivariateregression, we established that the skidding distan-ce, average bunching distance, and the number oflogs in the load (Table 4) simultaneously influence



Fig. 3 Cutters assistance according to skidding distance by groups

Slika 3. Ovisnost pomo}i sjeka~a o udaljenosti privla~enja po skupinama

Table 4 Coefficients and correlations of multiple linear regression

Tablica 4. Koeficijenti i korelacije multivarjantne linearne regresije

Factors

Faktori

Coefficients – Koeficijenti

tSignificant

Zna~ajnost

Partialcorrelation

Parcijalnakorelacija

BStd. Error

St. pogre{kaBeta

Constant – Konstanta –3.038 1.272 – –2.388 0.019 –

Skidding distance – Udaljenost privla~enja drva 0.004 0.002 0.233 2.924 0.004 0.286

Bunching distance – Udaljenost skupljanja drva 0.138 0.056 0.193 2.452 0.016 0.243

cutters assistance. With every meter of timber skid-ding, the cutters assistance increases by 0.004 mi-nute, with every meter of bunching distance by 0.139minute and with every additional log in the load by0.780 minute. Small differences between partial regres-sion coefficients, showing the correlation betweencutters assistance and working condition factor withconstant value of other two factors, and Pearsoncorrelation coefficients (Table 3) indicate that eachfactor directly influences cutters assistance. The num-ber of logs in the load accounts for 25% of totalvariability, the skidding distance 5% and the bunch-ing distance 4% (the square of part correlations).Multivariate model accounts for 40.7% of cuttersassistance variability (p <0.000).

Since we wanted to find out whether the workingmethod within the groups had any influence oncutters assistance, we excluded the influence of work-ing conditions (skidding and bunching distance, andthe number of logs in the load) from cutters as-sistance. By comparing the means of the residues bygroups after regression, we established that there areno statistically characteristic differences between thegroups in cutters assistance (Welch, p = 0.998), mean-ing that we explained all the differences in cuttersassistance, established at the beginning of this chap-ter, with the differences in working conditions.

7. Discussion – Rasprava

With this research we proved that working con-ditions determined by skidding distance, bunchingdistance and number of logs in the load influence thecutters participation in timber skidding in groupwork. There are no differences in work methodsbetween work groups that would reflect themselvesin absolute values of cutters assistance. The resultdoes not negate differences in the group work met-hods, established by Klun and Poje (2000a, 2000b)with the comparison of cutters assistance frequencywithin individual operation, because the compari-son of absolute values provides us with a generalanswer to the assistance increase with changing ofworking conditions and other conditions, while thefrequency distribution of assistance within groupsanswers to the differences of the method for achiev-ing maximal group effect.

The research results can be applied to loggingorganization, especially in work arrangement andeducation of workers, and directly in measurementand defining the standards for group work. Whendefining organizational form for cutting and skidd-ing in group work we also need to take into consi-deration working conditions, i.e. skidding distance,bunching distance and the number of logs in the

load. By increasing the value of the above factors, thetimber skidding becomes a bottleneck for total effectof the group, leading to greater need for additionalskidding capacities or inversely, decreasing the needfor additional cutters. Also Ze~i} and Krpan (2004),and Ze~i} and Maren~e (2005) came to similar con-clusions, but mainly in regard to skidding distance.When determining the group size, that has the com-mon standard determined on the basis of daily tim-ber quantity on the roadside landing, the basic prin-ciple of maximal means utilization for timber skidd-ing has to be followed. On the other hand, it meansthat a cutter or two cutters should also participate inskidding process in good working conditions fortimber skidding, which is frequently impossible, dueto great skidding effectiveness (not utilization). Withbetter working conditions for timber skidding, thecutters cut the pre-determined quantity of timberwith great difficulty. For adjustment of cutting andskidding processes in practice, the sum of cuttingdaily standard times in all working conditions has tobe lower than with timber skidding process. Forincreasing group efficiency and productivity, it isnecessary to provide detailed planning of workingsites and adjustable size of working groups. Besidepositive effects, these arrangements can also havenegative consequences, especially in the increasedeffect control and more precise work planning. Forsuccessful introduction of the changed working met-hod, it is necessary to achieve positive attitude ofworkers to planned changes.

Since cutters assistance between groups accord-ing to working conditions is not different, cutterswithin groups participate in timber skidding processin the same way, although they are not additionallyeducated for group work. It is obvious that the mo-tive for bigger salaries is by itself a motivation bigenough to demand rational behavior of workers ac-cording to working conditions. However, the work-ers should be informed about the risks related togroup work. The greater number of workers on thesame site and intertwining processes of cutting andskidding demands the increased and constant workand coworker movement control, and also the needfor worker qualification for cutting and skidding.Special attention should also be paid to a decrease ofwork difficulty, especially in extreme working con-ditions in skidding process (i.e. good and bad work-ing conditions). Both instances can cause »burnout«of workers because of cutting and skidding non-ad-justment.

Classic division of cutting and skidding can leadus to false and misleading conclusions when studyinggroup work efficiency. Due to overlap of cutting andskidding, apart from traditional working operations



the new arise, whereas synergetic effects prevent thesimple sum of standard times. The same holds truefor the division of productive and unproductive timewithin working processes. For example, from thecutters standpoint the participation in the skiddingprocess is regarded as unproductive time, or delaybecause of organization, whereas from the groupwork standpoint, it is regarded as productive time,necessary for effect increase of group work.

The ergonomic and economic development ofgroup work with existing »traditional« work tech-nology still has its potential. Apart from flexiblework groups adjustable to working conditions ac-cording to size (Ze~i} and Maren~e 2005) and capa-bility of various forest work performance (Winkler1990), it is reasonable to at least try to follow therecommendations for workers participation in plan-ning process and decision making, and constant edu-cation (Krivec 1979, Lewark et al. 1996). This bringsbetter relationships between group members andbetween group and company, and also increases theinformation flow and innovation possibilities.

8. Conclusion – Zaklju~ak

Group work is still one of the most perspectiveorganizational forms in forest work. This is not onlytrue for »traditional« form where chainsaws anddifferent kinds of tractors are used, but also for high-ly mechanized technologies with harvesters and for-warders. The aim of group work is to increase workefficiency by simultaneously decreasing, or at leastpreserving the level of work difficulty.

The simultaneous cutting and skidding proces-ses, and the common group daily standard causecutting and skidding overlap. On one hand this de-pends on cutters and capacity of skidding means,and on the other hand on working and environ-mental conditions and specific method of group work.In this research of group work we focused on theinfluence of working conditions on cutters participa-tion in timber skidding process.

The research took place in decreased range on 5places of research in Slovenian state forests, wherewe conducted time studies of timber skidding withIWAFUJI–T41 tractor in 3 working groups, with twocutters and one tractor driver. In our analysis weincluded 100 cycles of downhill timber skidding.

With multivariate analyses we confirmed the hy-pothesis that three working condition factors simul-taneously influence the duration of cutters parti-cipation in timber skidding process. Thus with everymeter of timber skidding, the cutters assistance isincreased by 0.004 minute, with every meter ofbunching distance for 0.139 minute and with every

additional log in the load by 0.780 minute. The num-ber of logs in the load has the greatest influence on theassistance duration. Multivariate model accounts for40.7% of variability. After eliminating the workingcondition influence on cutters assistance, we dis-proved the hypothesis that the groups have a dura-tion-specific work method, as other research projectshave indirectly indicated.

The results can be applied to define the workinggroup size that depends on all three factors at thesame time and not only on the skidding distance asthe other researchers suggested. By increasing allthree factors, the skidding means becomes a bottle-neck for the whole group effectiveness, which basi-cally means that we have to increase the timberskidding capacity, or decrease the number of cutters.The size flexibility of the working group brings theneed for more detailed work organization. Insigni-ficant differences between cutters assistance dura-tion in different groups mean that the group dailyefficiency is a motivational factor big enough for therational work. The result does not exclude educationand qualification of workers for safe group workand also the need for decreasing work difficulty.


Cimper{ek, M., 1987: Skupinsko delo v gozdni proizvod-nji (Group work in forestry production). Gozdarski vest-nik 45(6): 277–284.

Findeisen, E., 2002: ThüringenForst-Erfahrung zur teil-autonomen Gruppenarbeit in der hochmechanisierten Holz-ernte (ThuringenForst – experiences with partial auto-nomous group work for highly mechanized timber har-vesting). Forstechnische Informationen 4: 37–43.

Kerne`a, J., 1999: U~inki se~nje in spravila lesa pri Sne`nikd.d. po letu 1980 (Wood cutting and timber logging pro-ductivity in joint stock company Sne`nik by the year 1980).BSc. Thesis, University of Ljubljana – Biotechnical Faculty,Department of Forestry and Forest Resources, p. 30.

Klun, J., Poje A., 2000a: Spravilo lesa z zgibnim traktorjemIWAFUJI T–41 in po{kodbe pri se~nji in spravilu (Timberskidding with IWAFUJI T–41 skidder and stand damagedue to cutting and logging operations). BSc. Thesis, Uni-versity of Ljubljana – Biotechnical Faculty, Department ofForestry and Forest Resources, p. 150.

Klun, J., Poje A., 2000b: Spravilo lesa z zgibnim traktorjemIWAFUJI T–41 (Timber skidding with IWAFUJI T–41 skid-der). Gozdarski vestnik 58(7–8): 291–303.

Korbar, U., 1988: Skupinsko delo v gozdni proizvodnji(Group work in forestry production). BSc. Thesis, Univer-sity of Ljubljana – Biotechnical Faculty, Department ofForestry and Forest Resources, p. 49.



Krivec, A., 1979: U~inkovitost in oblikovanje novih organi-zacijskih postopkov pri spravljanju lesa s traktorji (Effi-ciency and modeling of new organizational methods bywood extraction with tractors). Gozdarski vestnik 37(7–8):315–323.

Kri`, A., 1984: Izku{nje skupinskega dela v gozdarstvu naposestvu Sne`nik, Ko~evska Reka (Experience with workgroup in forestry in company Posestvo Sne`nik, Ko~evskaReka). Gozdarski vestnik 42(6): 266–270.

Kruh, A., 1989: Posami~no in skupinsko delo v gozdniproizvodnji (Individual and group work in forestry pro-duction). BSc. Thesis, University of Ljubljana – Biotechni-cal Faculty, Department of Forestry and Forest Resources,p. 35.

Lewark, S., Strömquist, L., Kastenholz, E., Meier, D., 1996:Mit teilautonomen Gruppen zu höherer Effizienz undQualität der Waldarbeit? (Semi-autonomous groups forhigher efficiency and quality in forest work). Forstech-nische Informationen 11: 109–115.

Winkler, I., 1990: Skupinsko delo v gozdni proizvodnji(Group work in forestry production). Zbornik gozdarstvain lesarstva 35: 69–82.

Ze~i}, @., Krpan, P. B., 2004: Efficiency of group work inharvesting mountainous broadleaf thinning stand. Zbor-nik gozdarstva in lesarstva 74: 41–58.

Ze~i}, @., Maren~e, J., 2005: Mathematical models for opti-mization of group work in harvesting operation. CroatianJournal of Forest Engineering 26(1): 29–37.

Saètak

Utjecaj radnih uvjeta na preklapanje sje~e i privla~enja drvapri skupnom radu

Klasi~no pridobivanje drva u kojem se sje~a i sortimentnom metodom izradba drva obavljaju ru~no-strojnimradom uz primjenu motornih pila lan~anica, a privla~enje drva skiderom, zadràt }e, kao na~in rada, svoju va`nosti u budu}nosti, posebno u {umama koje se odlikuju te{kim terenskim radnim uvjetima te u manjim {umama privat-nih {umoposjednika. Osnovne zna~ajke skupnoga rada, kao jednoga od organizacijskoga oblika pridobivanja drva,jesu rad u skupini najmanje dvaju radnika koji istodobno obavljaju sje~u i izradbu te privla~enje drva te pri tomepomaù jedan drugomu. Opisani je na~in rada preteìt organizacijski oblik pridobivanja drva u slovenskimdràvnim {umama.

Veli~ina radne skupine ovisi ponajvi{e o kapacitetu stroja koji se koristi pri privla~enju drva.

Radnici u radnoj skupini imaju jedinstven (o~ekivan) dnevni u~inak te svi primaju jednaku naknadu za radzasnovanu na ostvarenim rezultatima. Na taj su na~in radnici motivirani sudjelovati u onim fazama rada koje sepokaù kao »usko grlo« pri ostvarivanju optimalnoga dnevnoga u~inka.

U slovenskom je {umarstvu skupni rad u proces pridobivanja drva uveden ranih 70-ih godina pro{loga stolje}a.Skupni rad zahtijeva svestranoga radnika koji je sposoban obaviti bilo koji posao u svojoj radnoj skupini. Od 1979.godine mnogi su se autori, u Sloveniji i u drugim europskim dràvama bavili, bilo na teoretskim postavkama biloprakti~nim istraìvanjima, skupnim radom (Kri` 1984, Cimper{ek 1987, Korbar 1988, Kruh 1989, Winkler 1990,Lewark i dr. 1996, Kerneà 1999, Klun i Poje 2000a, Klun i Poje 2000b, Findeisen 2002, Ze~i} i Maren~e 2005 idrugi).

Pri skupnom radu, kao organizacijskom obliku pridobivanja drva, za razliku od individualnoga rada, preklapa-ju se sje~a i izradba s privla~enjem drva. Uz zahtjev sigurnosti i ergonomske orijentiranosti, za skupni je rad po-sebno va`no postizanje najve}e mogu}e u~inkovitosti. To zna~i (slika 1) da je zbroj vremena sje~e i izradbe drva,privla~enja drva i vremena kada dolazi do preklapanja tih dviju faza rada minimalan. Navedena vremena ovise oradnim uvjetima u sje~ini te primijenjenom postupku rada (samo je dio vremena privla~enja drva: vo`njaneoptere}enoga zglobnoga traktora, vo`nja optere}enoga zglobnoga traktora i privitlavanje, neovisan o organiza-cijskom obliku skupnoga rada, a ponajvi{e pod utjecajem tehni~kih zna~ajki primijenjenoga zglobnoga traktora).

Kada su na~in rada, organizacija rada i tip zglobnoga traktora odre|eni i stalni, tada na vrijeme preklapanjaobaju radnih procesa utje~u radni uvjeti koji se javljaju pri sje~i i izradbi te pri privla~enju drva. Lo{i radni uvjetipri sje~i i izradbi drva (tanka stabla, gusta sastojina) op}enito smanjuju udio sjeka~a pri privla~enju drva te pove-}avaju udio traktorista u sje~i i izradbi drva; vrijedi i obrnuto kada su uvjeti privla~enja drva lo{i (velika udaljenostprivla~enja, mala sje~na gusto}a, strm teren, veliki uzdu`ni nagib traktorskih putova).

Ovim je istraìvanjem dan prilog dosada relativno slabo istraènim odnosima izme|u sje~e, izradbe i privla~e-nja drva pri skupnom radu, osvijetljen je problem promjenjivih radnih uvjeta u {umi (od sje~ine do sje~ine) uz uva-



àvanje ergonomskih, ekonomskih i organizacijskih ~imbenika. Istraìvanja su ograni~ena na udio sjeka~a u privla-~enju drva. Postavljene su dvije hipoteze:

Þ îmbenici radnih uvjeta (udaljenost privla~enja, udaljenost privitlavanja, premje{tanje zglobnoga trakto-ra pri oblikovanju tovara, broj komada trupaca u tovaru i obujam tovara) pove}avaju vrijeme preklapanjasje~e i izradbe s privla~enjem drva (pomo} sjeka~a traktoristu).

Þ Vrijeme trajanja pomo}i sjeka~a traktoristu razli~ito je od radne skupine do radne skupine (bez obzira na to{to su razli~iti radni uvjeti uzeti u obra~un).

Potvrda prve hipoteze omogu}ila je optimizaciju radne skupine sukladno u~inkovitosti rada i teìni radnihuvjeta. Pozitivan odgovor na drugu hipotezu upu}uje na potrebu izobrazbe radnika u skupni te isti~e va`nostdnevnoga u~inka skupine kao motivacijskoga ~imbenika.

Istraìvanja su provedena u devet odjela u slovenskim dràvnim {umama na nadmorskim visinama od 470 do1070 m. Osnovni su podaci o istraìvanim odsjecima prikazani u tablici 1. U odjelima 1 – 7 provedena je redovita, au odjelima 8 i 9 sanitarna sje~a stabala. Pra}en je rad {est skupina radnika, a svaka se radna skupina sastojala odjednoga traktorista i dvaju sjeka~a (1 + 2). Za privla~enje je kori{ten zglobni traktor IWAFUJI–T41 bez daljinskiupravljanoga vitla. Najve}a prosje~na dobna razlika izme|u radnih skupina bila je 8,7 godina, a najve}e prosje~noradno iskustvo izme|u radnih skupina 13,5 godina.

U pravilu su glavninu sje~e i izradbe obavljali sami sjeka~i, tek kod specifi~nih i najzahtjevnijih radnih opera-cija ili radnih operacija propisanih zakonom (npr. uhrpavanje grana), traktorist je pomagao sjeka~ima. Ve}inu jeradnih operacija privla~enja drva obavljao traktorist, sjeka~i su mu pomagali pri oblikovanju tovara. Povremeno jedodatno prerezivao trupce traktorist na pomo}nom stovari{tu. Dnevni u~inak skupine predstavlja izra|eno drvona pomo}nom stovari{tu.

Terenska su mjerenja obavljana 29 dana tijekom ljetnih mjeseci 1996. i 1997. godine. Na udaljenosti je pri-vla~enja od 1000 m snimljeno 218 turnusa privla~enja, od ~ega 78 uzbrdo, 114 nizbrdo te 26 na terenu bez utjecajanagiba (tablica 2). Prosje~an nagib traktorskoga puta odre|en je kao aritmeti~ka sredina pojedinih nagiba na tomtraktorskom putu, a duljine pojedinoga nagiba uzete su kao teìne. Traktorski put prosje~noga uzdu`noga nagibave}ega od +5 % predstavljao je privla~enje uzbrdo, traktorski put prosje~noga uzdu`noga nagiba manjega od –5 %predstavljao je privla~enje nizbrdo, a traktorski put prosje~noga uzdu`noga nagiba izme|u navedenih vrijednostiodredio je privla~enje drva po terenu bez utjecaja nagiba.

Zbog zna~ajnih razlika izme|u snimljenoga broja turnusa privla~enja drva izme|u radnih skupina (posebno sobzirom na smjer privla~enja drva: uzbrdo, nizbrdo, ravno) daljnja je analiza uklju~ila 100 turnusa privla~enjadrva nizbrdo, snimljenih u radnim skupinama 2, 3, 4 na mjestima istraìvanja 1, 3, 4, 5 i 6. Udaljenost privla~enjadrva mjerena je od mjesta kona~noga formiranja tovara do mjesta odvezivanja tovara (pomo}no stovari{te).Prosje~na udaljenost privitlavanja izra~unata je kao prosjek udaljenosti privitlavanja pojedinoga trupca u tovaru.Premje{tanje zglobnoga traktora pri formiranju tovara predstavlja zbroj svih pomicanja stroja u sastojini doformiranja punoga tovara. Obujam je tovara zbroj obujma svih trupaca u tovaru.

U tablici 3 prikazane su prosje~ne vrijednosti i korelacija izme|u trajanja vremena pomo}i sjeka~a i ~imbenika kojiodre|uju radne uvjete. Usporedbom prosje~noga trajanja vremena pomo}i sjeka~a izme|u radnih skupina uo~avamorazlike (p <0,000), a primjenom razlikovnoga testa sredina (Tukey HSD) dokazano je kako je prosje~no vrijemetrajanja pomo}i sjeka~a statisti~ki zna~ajno izme|u radnih skupina 2 i 3 te 2 i 4, dok nema statisti~ke potvrde razlikeizme|u radnih skupina 3 i 4. Prema prikazu na slici 2 moè se zaklju~iti da vrijeme pomo}i sjeka~a traktoristu raste spove}anjem udaljenosti privla~enja drva, odnosno s velikom vjerojatno{}u moèmo tvrditi da na vrijeme pomo}isjeka~a traktoristu osim metode rada i organizacije pojedine skupine utje~e i duljina udaljenosti privla~enja, {to sepodudara s teoretskim postavkama i rezultatima istraìvanja drugih autora (npr. Ze~i} i Maren~e 2005).

Prou~avanjem linearne ovisnosti izme|u individualnih utjecajnih ~imbenika radnih uvjeta i trajanja vremenapomo}i sjeka~a otkrili smo da duljina vremena pomo}i sjeka~a raste s porastom svih utjecajnih ~imbenika, asukladno pozitivnoj vrijednosti koeficijenta korelacije (r). Korelacija je statisti~ki ispitana za udaljenost privla~e-nja, udaljenost privitlavanja i broj komada trupaca u tovaru. Vrijednosti koeficijenata korelacije relativno su niske,{to indicira slabu korelaciju izme|u trajanja vremena pomo}i sjeka~a i individualnih utjecajnih ~imbenika radnihuvjeta. Primjenom linearne multivarjantne regresije (tablica 4) èlio se ispitati individualni utjecaj pojedinoga~imbenika radnih uvjeta (uz stalne vrijednosti ostalih ~imbenika radnih uvjeta) na trajanje vremena pomo}isjeka~a. Utvr|eno je da udaljenost privla~enja, udaljenost privitlavanja i broj komada trupaca u tovaru istodobnoutje~u na pomo} sjeka~a.

Sa svakim metrom privla~enja trajanje vremena pomo}i sjeka~a raste 0,004 min., sa svakim metrom privitla-vanja trajanje vremena pomo}i sjeka~a raste 0,139 min., a sa svakim dodatnim trupcem u tovaru za 0,780 min.



Broj komada trupaca u tovaru pokazuje varijabilnost od 25 %, udaljenost privla~enja 5 %, a udaljenostprivitlavanja 4 %. Multivarjantni model ima varijabilnost trajanja vremena pomo}i sjeka~a od 40 do 7 %.Statisti~ki je dokazano (Welchovim testom, p = 0,998) da na trajanje vremena pomo}i sjeka~a ne utje~e radnametoda unutar pojedine skupine ve} samo radni uvjeti.

Rezultati provedenoga istra`ivanja primjenjivi su pri organizaciji skupnoga rada u pridobivanju drva, u izo-brazbi radnika te za mjerenje i odre|ivanje u~inkovitosti skupnoga rada. Pri organizaciji skupnoga rada posebnupa`nju treba usmjeriti odre|ivanju radnih uvjeta u sje~ini imaju}i na umu ove ~imbenike: udaljenost privla~enja,udaljenost privitlavanja i broj komada trupaca u tovaru. Pove}avanjem vrijednosti nabrojenih ~imbenika pri-vla~enje drva postaje »usko grlo« pri ostvaraju propisanih dnevnih u~inaka skupine. Navedeni je problem mogu}erije{iti pove}anjem kapaciteta privla~enja drva ili suprotno, smanjenjem broja sjeka~a. Do sli~nih su zaklju~aka, aliprete`no u svezi s udaljeno{}u privla~enja drva, do{li Ze~i} i Krpan (2004) i Ze~i} i Maren~e (2005).

Pri optimizaciji radne skupine radi ostvarivanja njezina propisanoga dnevnoga u~inka nastoji se posti}inajve}a mogu}a u~inkovitost zglobnoga traktora. Za pove}anje u~inkovitosti radne skupine nu`no je detaljnoplaniranje radova. Skupni rad sa stajali{ta sigurnosti pri {umskom radu nosi odre|ene negativnosti i opasnosti.Radnici moraju biti osposobljeni za obje faze rada: sje~u i izradbu te privla~enje drva, preklapanje sje~e i izradbe sprivla~enjem drva tra`i pove}an i stalan rad, potreban je poseban oprez zbog ve}ega broja radnika i stroja/strojevakoji se kre}u na malom podru~ju i dr.

Treba povesti ra~una o ergonomskom i ekonomskom razvoju skupnoga rada. Preporu~uje se uklju~ivanjeradnika u postupak planiranja i dono{enja odluka te njihova stalna izobrazba (Krivec 1979, Lewark 1996). Tako }eodnosi izme|u radnika u radnoj skupini, ali i odnosi izme|u radnika i poslodavca postati bolji i kvalitetniji, ubrzat}e se protok informacija i pove}ati mogu}nost inovacija i unapre|enja.

Skupni je rad, uistinu, jedna od najperspektivnijih organizacijskih oblika {umskoga rada, i to ne samo za»klasi~no« pridobivanje drva gdje se sje~a i izradba obavlja ru~no-strojnim radom uz primjenu motornih pilalan~anica, a privla~enje drva razli~itim tipovima zglobnih traktora, ve} i za visokomehanizirano pridobivanje drvauz uporabu harvestera i forvardera.

Klju~ne rije~i: skupni rad, sje~a, privla~enje, radni uvjeti, multivarjantna analiza



Authors’ address – Adresa autorâ:

Anton Poje, MSc.e-mail: [email protected]. Prof. Igor Poto~nik, Ph.D.e-mail: [email protected] of Ljubljana, Biotechnical FacultyDepartment of Forestry and Forest ResourcesVe~na pot 831000 LjubljanaSLOVENIA

Received (Primljeno): November 15, 2007Accepted (Prihva}eno): December 6, 2007

Forwarding productivity in SouthernAustria

Mohammad Reza Ghaffarian, Karl Stampfer, John Sessions


Forwarders are common machines in the cut to length harvesting system. A generalregression model for predicting the time of forwarding was developed using 82 workingcycles of two kinds of machines in downhill timber extraction. Using stepwise regression,variables such as forwarding distance, slope, type of forwarder and piece volume weresignificantly entered to the model. Increasing forwarding distance increases the forwardingtime. However, if piece volume and slope of the trail increase, the forwarding time decreases.The average forwarding production was estimated to about 17.9 m3/PSH0 (ProductiveSystem Hours) while mean load per trip was 10.04 m3 and average forwarding distance was97 m.

Keyword: forwarder, productivity, cost, model, regression


Cut-to-length systems generally comprise twomachines: a harvester and a forwarder. In Aus-tria, 17% of the timber is extracted by forwarders(www.lebensministerum.at). Many factors canaffect the productivity of forwarders. Kellogg andBettinger (1994) in Oregon (USA) developed a pro-ductivity model for forwarding, which included vari-ables such as assortments (pulpwood, sawlog ormixed loads), volume per load, travel distance ofunloaded forwarder, travel distance during loadingand travel distance of loaded forwarder to landing.The productivity of forwarder is strongly correlated tostand type, average extraction distance, timber vo-lume density at the strip road and load volume of theforwarder (Tufts and Brinkeer 1993, Kuitto et al. 1994).

In other studies carried out in central Finland fordifferent cutting sites, harvesting density on striproads, the average extraction distance, forwarderload capacity, timber assortment and bunching ofassortments on the strip road had an effect on thehaulage system (Nurminen et al. 2006).

The working method can also affect the producti-vity of harwarders as well as variables such as load

volume, tree size and extraction distance (Anders-son and Eliasson 2004).

A study of short-wood forwarding carried outin Northern Spain resulted in a productivity of6 to 15 t/SMH (Spinelli et al. 2003). Akay andSessions (2001) reported the production rates in USAfor small, medium and large size of forwarder as0.51, 0.6 and 0.69 m3/h, respectively.

Other productivity studies resulted in the rangeof 8 to over 20 t/SMH, depending on the model andworking conditions (UK Forestry Commission 1995,Gullberg 1997, Martin dos Santos et al. 1995, Saun-ders 1996, Goglia et al. 1999, Horvat et al. 1990).

Cordeo et al. (2006) used GPS technology to mo-nitor cut-to-length (CTL) system in Chile. In thismethod, by capturing the position of machine, it waspossible to generate surface progress grids, whichwhen combined with inventory grids, result in yieldinformation by surface and time unit. The hourlyproduction rate was 35.8 m3 in clear cut sites.

Due to increasing forwarding system in the forestsof central Europe, it is necessary to study the pro-ductivity and cost of this system to give planners auseful tool for harvest planning. There is no generaltime predicting model available for forwarders inAustria. Therefor this study was carried out to de-



velop such a model using time study databases ontwo types of forwarders. Verification of the modelwas done. The effect of variables on the time offorwarding is also presented in this paper.

2. Method of study – Metoda istraìvanja

2.1 Site of study – Mjesto istraìvanja

The first study was carried out in Weiz in Steier-mark in southern Austria. In this area a Ponsse Buf-falo Dual forwarder was used in a mixed stand ofspruce, fir, larch and pine. The terrain slope was mo-derate (11%).

By the Dual concept of a combined harvester-for-warder system, known as the harwarder, trees arefelled, delimbed, topped, and bunched using theharvesting head of the harwarder in pre-plannedmachine trails. The operator then fits the load bunkfor forwarding the logs to the roadside landing. Thegrapple head is used to load the logs onto the bunkof the machine. For the machine specifications of theharwarder refer to Table 1.


M. REZA GHAFFARIAN et al. Forwarding productivity in Southern Austria (169–175)

Fig. 1 Ponsse Buffalo Dual Harwarder

Slika 1. Harvarder Ponsse Buffalo Dual

Table 2 Specifications for the Ponsse Buffalo Dual Harwarder

Tablica 2. Tehni~ke zna~ajke harvardera Ponsse Buffalo Dual

Vehicle technical data – Tehni~ki podaci vozila

Engine – Motor Mercedes-Benz OM906LA

Power – Snaga 180 kW

Torque – Moment 900 Nm @ 1400 min-1

Pump capacity – Obujam pumpe 100 cm3

Fuel tank volume

Obujam spremnika goriva130 L

Length – harvester

Duljina kad se koristi kao harvester8850–9150 mm

Length – forwarder

Duljina kad se koristi kao forvarder9400–9950 mm

Width (600/700 tires)

[irina (gume 600/700)2670–2810 mm

Ground clearance – Klirens 690 mm

Tare mass of harvester

Masa praznoga harvestera15700 kg

Tare mass of forwarder

Masa praznoga forvardera16400 kg

Load capacity – Nosivost 14000 kg

Length of loading space

Duljina tovarnoga prostora4040–4590 mm

Extension of loading space

Produljenje tovarnoga prostora0–700 mm

Harvesting head – Sje~na glava

Harvesting head – Harvesterska glava Ponsse H53


Length – Duljina 64 cm

Felling diameter – Sje~ni promjer 52 cm

Feed force – Posmi~na sila 18 kN

Feed speed – Posmi~na brzina 4 m/s

Number of delimbing knives

Broj noèva za kresanje grana5

Measuring system – Mjerni sustav Ponsse Opti 4G

Table 1 Description of study sites

Tablica 1. Opis mjesta istraìvanja


Area, ha – Povr{ina, ha 2.27 1.83

Slope, % – Nagib terena, % 11 39

Stand age, years – Dob sastojine, godine 70–130 90

No. of trees per hectare – Broj stabala po ha 1089 729

Growing stock, m3/ha – Drvna zaliha, m3/ha 510.4 646

No. of harvested trees – Br. posje~enih stabala 1073 470

Harvested volume, m3 – Posje~eno drvo, m3 331.8 513

Tree volume, m3 – Srednji obujam stabla, m3 0.31 0.7

Harvesting intensity, % – Intenzitet sje~e, % 28.7 45

Number of trails – Broj traktorskih vlaka 15 5

Length of trails, m – Duljina vlaka, m 40–200 190–235

Time of harvesting – Vrijeme sje~e Spring – Prolje}e

The second site of study was near to Muerzzu-schlag in Steiermark. The mixed species stand con-sisted of spruce, larch and fir. A Gremo 950R for-warder was used to extract logs that had been pro-cessed and piled by a harvester. The terrain slopewas 39%. In steep terrain, this forwarder would usea cable fixed to a standing tree to allow safe opera-tion on steep trails (Wratschko 2006).

The general stand and terrain characteristics onboth sites are presented in Table 1. According toinformation from the forest office in Steirmark, theharvesting density was 100 m3/ha with a mean DBHof 25 cm.

2.2 Data collection – Prikupljanje podataka

A continuous time study method was used inboth production studies using an electronic EG20timer. A typical work cycle included loading, travelloaded, unloading and travel empty.

Loading element included the time used to loadthe logs for one trip. Travel loaded was defined asthe time to move a loaded machine to the landing.Unloading included the time spent to unload thelogs on the landing and travel empty consisted in thetime to move empty from the landing to loading site.The delays of more or less 15 minutes were recordedas well as miscellaneous delays in each workingcycles.

The same variables were used at both study sites.Forwarding distance, piece volume, total load vo-lume and slope were recorded during the datacollection. Data for 82 working cycles were collected.

It was assumed that forwarding time was a func-tion of forwarding distance, piece volume, slope oftrail and type of forwarder. Two time databases wereused to develop the time prediction model usingstepwise multiple regression.

3. Research results – Rezultatiistraìvanja

3.1 Productivity – Proizvodnost

The observed productivity was 17.9 m3/PSH0 andthe average load per cycle was 10.04 m3. The averageloaded and empty travel speeds of the forwarder were35.9 m/min (0.6 m/s) and 26.3 m/min (0.45 m/s),respectively. Loaded travel was in a downhill direc-tion which resulted in a higher speed than travelempty.

Using the system cost for the Buffalo Dual Har-warder of 120 EUR/h (Affenzeller 2005), the averageforwarding cost was estimated to 6.72 EUR/m3.

3.2 Delays – Prekidi

The percentage delay times and time for fixingand opening the cable, relative to total time ofworking for each forwarder type is presented inTable 5.

The total number of delays was greater for theGremo 950R forwarder than for Ponsse Buffalo DualHarwarder. The Gremo 950R forwarder had greaternumber of delays exceeding 15 minutes, whilePonsse Buffalo Dual Harwarder had greater numberof short delays.

3.3 Model – Model

SPSS software was used for processing the step-wise multiple regression. Stepwise multiple regres-


Forwarding productivity in Southern Austria (169–175) M. REZA GHAFFARIAN et al.

Table 3 Specification of Gremo 950R Forwarder

Tablica 3. Tehni~ke zna~ajke forvardera Gremo 950R

Engine – MotorDEUTZ Type BF4M

1013 EC


Pump capacity – Obujam pumpe 100 cm3

Hydraulic tank volume – Obujam spremnika ulja 100 L

Fuel tank volume – Obujam spremnika goriva 110 L

Tyre dimension – Dimenzije guma 600/50–22.5

Load capacity – Nosivost 10000 kg

Boom – Dizalica Loglift type 51F

Range – Doseg 6.5 m

Lifting capacity – Podizni moment 59 kNm

Length – Duljina 7895 mm

Width – [irina 2600 mm

Height – Visina 3445 mm

Turn radius – Polumjer okretanja 12.7 m

Ground Clearance – Klirens 580 mm

Tare mass of forwarder – Masa praznoga forvardera 11185 kg

Table 4 Worktime delays

Tablica 4. Prekidi rada

Forwarder – Forvarder Gremo Ponsse

Delays – Prekidi (<15min) 5.87 % 10.6 %

Delays – Prekidi (>15min) 11.87 % 5.49 %

Miscellaneous delays – Slu~ajni prekidi – 0.51 %

Total delays – Prekidi ukupno 17.74 % 16.6 %

Fixing the cable – Postavljanje uèta 0.25 % –

Opening the cable – Odvezivanje uèta 0.85 % –

sion assumes that if any variable has a significanteffect on the Residual Mean Squares (RMS) of themodel, it will be included in the model. Forwardingdistance, piece volume, type of machine and slopewere all significant variables at a = 0.05.

The following forwarding model was determinedon the basis of 82 recorded working cycles:

t = 81.293 – 47.886 × V – 46.795 × F ++ 0.076 × L – 1.189 × I

where:t – forwarding time, min/cycleV – piece volume, m3/pcs.F – type of machineL – forwarding distance, mI – slope (inclination) of skid trail, %

The value for the Ponsse Buffalo Dual Harwarderis 1 and the value of 0 is considered for Gremo 950Rforwarder. The multiple correlation coefficient (R2)of 0.32 is interpreted as 32% of total variability, which



Table 5 ANOVA table – Overall goodness of fit

Tablica 5. Analiza varijance – Testiranje zna~ajnosti modela

Sum of Squares

Zbroj kvadrata

Degree of freedom

Br. stupnjeva slobode

Mean Square

VarijancaF

Significance

Zna~ajnost

Regression model – Regresijski model 2805.529 4 701.382 8.817 .000

Residual – Ostatak 5966.085 75 79.548

Total – Ukupno 8771.614 79

Fig. 2 Cycle time vs. forwarding distance

Slika 2. Ovisnost vremena turnusa o udaljenosti izvo`enja

Fig. 3 Cycle time vs. skid trail slope

Slika 3. Ovisnost vremena turnusa o nagibu vlake

Fig. 4 Cycle time vs. piece volume

Slika 4. Ovisnost vremena turnusa o obujmu komada

is explained by the regression equation. The signi-ficant level of ANOVA (0.000) shows that the model

is significant at a = 0.05 (Table 5).

The effect of each variable on forwarding timewas studied by changing one variable while holdingthe other variables constant at their mean value.Forwarding time includes travel empty, loading, tra-vel loaded, and unloading.

Figures 2, 3, 4 and 5 show the effect of forwardingdistance, slope of skid trail, piece volume and type ofmachine on forwarding time respectively within therecorded range of variables. In Figure 6, the percent

of time spent on each element of the forwardingcycle is shown. The most time is spent loading thelogs. Table 6 presents the summary statistics.


The variables such as forwarding distance, piecevolume, type of machine and slope of the skid trailwere entered into the general model for predictingforwarding time as significant variables, which canbe applied in logging planning. Increasing forward-ing distance will increase forwarding time, but ifpiece volume and downhill slope increase, the for-warding time decreases.



Fig. 5 Cycle time vs. type of machine

Slika 5. Ovisnost vremena turnusa o tipu vozila

Fig. 6 Percentage of each work component in the cycle time

Slika 6. Udjeli pojedinih sastavnica rada u vremenu turnusa

Table 6 Summary statistics of parameters

Tablica 6. Statisti~ki podaci istraìvanih parametara

Parameter

Parametar

Maximum

Najve}e opaànje

Mean

Aritmeti~ka sredina

Minimum

Najmanje opaànje

Travel Empty, min/cycle – Neoptere}eno kretanje vozila, min/tura 18.67 5.76 0.4

Loading, min/cycle – Utovar drva, min/tura 42.24 17.23 2.78

Travel Loaded, min/cycle – Vo`nja optere}enoga forvardera, min/tura 10.72 4.22 0.35

Unloading, min/cycle – Istovar drva, min/tura 15.31 6.5 0.97

Cycle time, min/cycle – Vrijeme turnusa, min/tura 57.68 33.72 8.9

Forwarding distance, m – Udaljenost izvoènja, m 280 96.64 4

Slope of skid trail, % – Nagib traktorske vlake, % 40 21.62 5

Load volume, m3/cycle – Obujam tovara, m3/tura 18.7 10.04 1.37

Piece volume, m3/pcs. – Obujam komada, m3/kom. 0.49 0.14 0.04

Follow-up research should examine the produc-tivity of uphill forwarding operations on a variety ofsites.


Affenzeller, G., 2005: Integrierte Harvester-Forwarder-Kon-zepte Harwarder. MSc. thesis, Institute of Forest Engineer-ing, University of Natural Resources and Applied LifeSciences Vienna.

Akay, A., Sessions, J., 2001: Minimizing road constructionplus forwarding costs under a maximum soil disturbanceconstraint. The International Mountain Logging and 11th

Pacific Northwest Skyline Symposium December 10–12,Seattle, Washington, USA, p. 268–279.

Andersson, J., Eliasson, L., 2004: Effect of three harvestingwork methods on Harwarder productivity in final felling.Silva Fennica 38(2): 195–202.

Cordero, R., Mardones, O., Marticorena, M., 2006: Evalua-tion of forestry machinery performance in harvesting ope-rations using GPS technology. Proceedings of IUFRO Pre-cision Forestry Symposium, Stellenbosch, 2006, p. 163–173.

Goglia, V., Horvat, D., Sever, S., 1999: Technical charac-teristics and test of the forwarder Valmet 860 equippedwith a Cranab 1200 crane. Forestry Faculty of Zagreb Uni-versity, Internal Report, p. 23.

Gulberg, T., 1997: Time consumption model of off-roadextraction of shortwood. Institutionen foer Skogsteknik,Sveriges Lantbruksuniversitet, Uppsatser och Resultat 297,p. 29.

Holzeinschlagmeldung, 2004: < http:// www.lebensmi-nisterium.at >

Horvat, D., Goglia, V., Sever, S., 1999: Technical charac-teristics and test of the forwarder Timberjack 1410 and

Timberjack 1710. Forestry Faculty of Zagreb University,p. 32.

Huggard, E. R., 1978: Optimum road spacing. Quarterlyjournal of forestry 72(4): 207–210.

Kellogg, D. L., Bettinger, P., 1994: Thinning productivityand cost for a mechanized cut-to-length system in theNorthwest Pacific coast region of the USA. J. For. Eng. 5(2):43–54.

Kuitto, P. J., Keskinen, S., Lindroos, J., Oijiala, T., Raja-maeki, J., Rasanen, T., Terava, J., 1994: Mechanized cuttingand forest haulage. Mestaeteho Report 410, p. 38.

Martin dos Santos, S., Machado, C., Leite, H., 1995: Techno-economical analysis of eucalyptus extraction with forwar-der in flat terrain. Revista Arvore, Vicosa 19(2): 213–227.

Nurminen, T., Heikki, K., Uusitalo, J., 2006: Time con-sumption analysis of the mechanized cut-to-length har-vesting system. Silva Fennica 40(2): 335–363.

Saunders, C., 1996: West Argyll Valmet 890 forwarder trial1996. Forestry Commission Research Division – TechnicalDevelopment Branch, Internet Project Information Note7/96: p. 9.

Spinelli, R., Owende, P., Ward, S., Torneo, M., 2003: Com-parison of short-wood forwarding systems used in Iberia.Silvia Fennica 38(1): 85–94.

Tufts, R. A., Brinker, R. W., 1993: Productivity of Scandi-navian cut-to-length system while second thinning pineplantations. Forest Product Journal 43 (11–12): 24–32.

UK Forestry Commision, 1995. Terrain classification. Tech-nical Note 16/95: p. 5

Wratschko, B., 2006: Production of cable forwarder. MSc.Thesis, Institute of Forest Engineering, University of Na-tural Resources and Applied Life Sciences Vienna.

Saètak

Proizvodnost izvoènja drva u ju`noj Austriji

Forvarderi se uobi~ajeno koriste pri sortimentnoj metodi izradbe drva. U Austriji se 17 % izra|enih drvnihsortimenata izvozi forvarderima. Mnogi ~imbenici utje~u na proizvodnost forvardera: vrsta drvnih sortimenata,obujam tovara, udaljenost izvoènja, sje~na gusto}a. Stoga su potrebna istraìvanja proizvodnosti i tro{kovasustava izvoènja drva koja bi bila upotrebljiva pri planiranju i organizaciji radova.

Istraìvanje je provedeno na dvije sje~ine na osnovi studija vremena pri radu dvaju razli~itih tipova strojeva:forvardera i harvardera. Cilj je istraìvanja razvoj modela proizvodnosti i tro{kova izvoènja drva s obzirom nautjecajne ~imbenike.

Na prvoj je sje~ini kori{ten harvarder Ponsse Buffalo Dual pri sje~i stabala i izvoènju drvnih sortimenata izmje{ovitih sastojina smreke, jele, ari{a i bora. Na drugoj je sje~ini kori{ten forvarder Gremo 950R za izvoènjedrvnih sortimenta, dok se za sje~u i izradu koristio harvester. Osnovne su zna~ajke sje~ina prikazane u tablici 1, a



{umskih vozila u tablicama 2 i 3. Prosje~ni je nagib terena na prvoj sje~ini iznosio 11 %, a na drugoj 39 %. Pri raduna strmom terenu forvarder je bio opremljen vitlom te se kretao od pomo}noga stovari{ta do mjesta utovara usmjeru uzdu`noga nagiba terena namatanjem uèta koje je bilo vezano na dube}e stablo.

Ukupno su snimljena 82 radna turnusa proto~nom metodom studija vremena. Radni turnus forvarderauklju~uje neoptere}eno kretanje vozila, utovar drva, optere}eno kretanje vozila te istovar drva. Prekidi su radazabiljeèni u svakom turnusu. Tako|er su mjereni utjecajni ~imbenici izvoènja drva: udaljenost izvoènja,obujam tovara, obujam pojedinoga drvnoga sortimenta te nagib traktorske vlake.

Ostvarena je prosje~na proizvodnost iznosila 17,9 m3 po pogonskom satu rada. Optere}eno se kretanje vozilaodvijalo niz nagib te su ostvarene ve}e brzine kretanja nego pri neoptere}enom kretanju vozila. Prosje~ni jejedini~ni tro{ak izvoènja drva iznosio 6,72 EUR/m3.

Linearnom multivarjantnom su se regresijom odredile statisti~ki zna~ajne varijable (udaljenost izvoènja,obujam komada, tip forvardera i nagib terena) te iskazao model utro{ka vremena pri izvoènju drva. U modeluvarijabla F iznosi 1 za Ponsse Buffalo Dual harvarder, odnosno 0 za Gremo 950R forvarder.

Utjecaj pojedinih ~imbenika na utro{ak vremena izvoènja drva prikazan je na slikama 2, 3, 4 i 5 u opseguizmjerenih vrijednosti. Slika 6 prikazuje postotni udio utro{ka vremena pojedine sastavnice radnoga turnusa, atablica 6 zbirne statisti~ke podatke mjerenih vrijednosti.

Rezultati istraìvanja pokazuju da se pove}anjem udaljenosti izvoènja pove}ava utro{ak vremena radaforvardera. Utro{ak se vremena izvoènja drva forvarderima smanjuje s pove}anjem obujma komada te pove}enjemnagiba vlake pri izvoènju drva nizbrdo. Daljnjim je istraìvanjima potrebno odrediti proizvodnost forvardera priizvoènju drva uz nagib pri razli~itim uvjetima rada.

Klju~ne rije~i: forvarder, proizvodnost, tro{ak, model, regresija




Mohammad Reza Ghaffarian, MSc.e-mail: [email protected]. Prof. Karl Stampfer, [email protected] of Natural Resources and Applied

Life Sciences ViennaDepartment of Forest and Soil SciencesInstitute of Forest EngineeringPeter Jordan Strasse 821190 ViennaAUSTRIA

Prof. John Sessions, PhD.e-mail: [email protected] State UniversityCollege of ForestryDepartment of Forest Engineering204 Peavy HallCorvallis, OR 97331-5706USA

Received (Primljeno): December 21, 2006Accepted (Prihva}eno): November 19, 2007

Costs and efficiency of timber harvestingby NIAB 5–15 processor mounted on a farm

tractor

Janusz Sowa, Dariusz Kulak, Grzegorz Szewczyk


The present research deals with costs and economic effectiveness of timber harvestingtechnology with the use of the NIAB 5–15 processor mounted on a farm tractor, whichmethod is used in Poland. Measurements were conducted in pine, fir and spruce stands,which underwent early and late thinnings. During harvesting, a time study was performedusing the continuous reading method. On completion of felling works, the volume of timberharvested was measured.

The efficiency and unit costs of timber harvesting were calculated in the operationalworking time for the chainsaw operator and processor operator.

In all analysed stands, significantly higher efficiency was observed in late thinnings than inearly ones. This resulted in higher economic effectiveness of this technology in the thinningsof older age classes. The approximation of regression functions allowed for the prediction offixed unit costs and efficiency depending on the average volume of trees being removed. Thelow share of fixed costs in the costs of exploitation by processor proves that a longer shift onlyaffects to a small degree the economic effectiveness of the analysed technology.

Keywords: timber harvesting, thinnings, processor, costs, productivity


Considering the fierce competition in the Polishmarket of forest services, timber harvesting com-panies can achieve the greatest advantage by usingwork methods of low unit costs. In Poland, the majo-rity of companies are small, with limited financialcapacity (Kocel 2003) and lack of highly efficient andvery expensive multi-operational machinery. The de-gree of harvesting mechanization may be increasedby means of machines accessible to middle-size forestservice companies thanks to the use of farm trac-tor-mounted processors (Karlsson 1988, Walczyk 1997).The use of such machines makes the working en-vironment relatively less strenuous (Giefing 1994b),which is another argument for their widespread in-troduction. However, in the free market economy, oneof the basic criteria of assessment of technologicalprocesses of timber harvesting is their economic ef-fectiveness. That is why it is necessary to analyse the

processor exploitation costs in detail before makinga decision to purchase it. At present there are noresults of such analyses conducted in Poland (Gief-ing 1994a), which may constitute an obstacle to theintroduction of processors in forest work.

The aim of the present research is to determinethe efficiency and timber harvesting costs by theNIAB 5–15 processor mounted on a farm tractor. Thescope of the research is limited to early and latethinnings, performed in pine, fir and spruce stands.

2. Research area – Podru~je istra`ivanja

The present research was located in southernPoland. Measurements were taken on square sam-ple plots with the surface of 0.5 ha each. The longerside of each plot ran along a skid trail. The basicfeatures of the researched stands and the characte-ristics of the measures taken in them are presentedin Table 1.



3. Used machines and harvestingtechnology – Istraìvani strojevi i na~ini

pridobivanja drva

The basic parameters of the NIAB 5–15 processormounted on a MTZ 100 farm tractor, used in theresearch, are presented in Table 2.

Harvesting was conducted according to the fol-lowing schedule. Using the chain saw, the operatorfelled the trees marked by the staff of the State ForestAdministration. The felling was performed in thedirection opposite to the skid trail. If a tree remainedsuspended, the operator cut only the hinge. Theprocessor operator stretched the cable and attachedit to the butt end of the tree. Then, he performedwinching operating the winch via radio, and some-times preceded by removing the suspension. At theskid trail, the operator detached the load and per-formed timber debranching and cross-cutting usingthe processor (Fig. 1). The timber was cut into 1.25-m


J. SOWA et al. Costs and efficiency of timber harvesting by NIAB 5–15 processor mounted on a farm tractor (177–184)

Table 1 Characteristics of researched stands

Tablica 1. Zna~ajke istraìvanih sastojina

Stand number

Broj sastojine1 2 3 4 5 6 7 8

No. of sample plots

Broj pokusnih ploha3 3 3 3 1 2 2 1

Category of utilisation

Vrsta prihoda

Early thinning

Rana proreda

Late thinning

Kasna proreda

Early thinning

Rana proreda

Late thinning

Kasna proreda

Early thinning

Rana proreda

Late thinning

Kasna proreda

Species composition, %

Omjer smjese, %Pinus – 100

Pinus – 90

Betula – 10

Abies – 70

Fagus – 30

Abies – 70

Fagus – 30

Picea – 70

Abies – 30

Picea – 50

Abies – 50

Picea – 80

Abies – 20

Picea – 90

Abies – 10

Age, years

Dob, godine25 45 37 52 35 40 80 60

Stand density index

Obrast0.9 0.7 0.9 0.6 0.8 1.1 1.0 1.2

Stand density

Sklop sastojine

full

potpun

moderate

djelomi~an

full

potpun

moderate

djelomi~an

moderate

djelomi~an

full

potpun

moderate

djelomi~an

moderate

djelomi~an

Stand quality class

Bonitetni razredIa

Ia

II I I I I I

Breast height diameter, cm

Prsni promjer, cmPinus – 13

Pinus – 22

Betula – 29

Abies – 13

Fagus – 14

Abies – 24

Fagus – 20

Picea – 14

Abies – 10

Picea – 13

Abies – 10

Picea – 35

Abies – 30

Picea – 24

Abies – 19

Height, m

Visina, mPinus – 12

Pinus – 20

Betula – 22

Abies – 12

Fagus – 12

Abies – 18

Fagus – 17

Picea – 14

Abies – 10

Picea – 16

Abies – 13

Picea – 30

Abies – 23

Picea – 23

Abies – 18

Growing stock, m3/ha

Drvna zaliha, m3/ha140 228 164 311 90 320 668 637

Mean cutting tree, m3

Srednje sje~no stablo, m30.06 0.32 0.07 0.38 0.09 0.09 0.56 0.47

Table 2 Technical data of the NIAB 5–15 processor

Tablica 2. Tehni~ki podaci procesora NIAB 5–15

Technical parameters – Tehni~ki podaci Values – Vrijednosti

Mass – Masa 1030 kg

Length – Duljina 2000 mm

Width – [irina 2450 mm

Height – Visina 2300 mm

Engine power – Snaga motora 30 kW

Maximum cutting diameter – Najve}i sje~ivi promjer 500 mm

Length of winch cable – Duljina uèta vitla 50 m

Cable winch pulling power – Vu~na sila vitla 25 kN

Working pressure – Radni tlak 20 MPa

Pump capacity – Protok ulja 60 dm3/min

Oil tank capacity – Obujam spremnika ulja 60 dm3

long pieces (Fig. 2). After the winching and timberprocessing was completed within the range of thecable, a new working point of the processor wasfound on the skid trail.

4. Methods of measurements andcalculations – Metode mjerenja i obrade

podataka

A time study of harvesting was performed usingdata loggers of the PSION type, equipped with ap-propriate software. The precision of time measure-ments was 1 s. The time study was conducted sepa-rately for the chainsaw operator and for the pro-cessor operator. The amount of the fuel used wasmeasured during each tank filling. After the comple-tion of harvesting, the volume of the timber har-vested was determined on the basis of log piecesdiameters.

It was assumed that the economic effectiveness ofthe analysed timber harvesting variants, performedduring thinnings, would be best characterized bytheir unit costs, expressed in �/m3. Considering themethods of cost calculation provided in the lite-rature (Suwa³a and Rzadkowski 2001, Suwa³a 1998,Zychowicz 1998), the following equipment harvest-ing costs (C) were taken into account:

C = Ca + Ci + Cfl + Cr [EUR/h] where:

Þ harvesting costs C [EUR/h]

Þ amortization Ca =P

T H��

��

��EUR

h

Þ interest of capital Ci =

P p

T H h2 100

�

��

��

��EUR

Þ fuel and lubrication cost Cfl [EUR/h]

Þ repair cost Cr =P

k

T A h

�

��

��

��100 EUR

Þ machine purchase price P [EUR]

Þ machine economic life T [years]

Þ scheduled operating time H [h/year]

Þ capital interest rate p [%]

Þ repair cost index n [%]

The calculations disregard the cost of wages formachine operators. According to other authors’work (Jod³owski 2000, Maciak and Skar¿yñski 1995,Porter 1998), all calculations were performed for theeffective work time. This allowed for the eliminationof contingency in the assessment of efficiency, caus-ed by e.g. the number of machine failures or thelength of breaks during work (Giefing and Gackow-ski 2001). Both the processor and the tractor underanalysis were in their post-amortization period. How-ever, for the sake of the calculations, it was assumedthat the machines were new (Tab. 3).

The cost of fuel and lubricants (Cfl) was deter-mined on the basis of their use in field conditions.

Unit costs c0 were calculated for each machine se-parately using the following formula:

c0 =C

V0

EUR

m3

�

��

��, where:

Þ V0 – productivity in the effective time (m3/h).

Unit costs for the whole technology were deter-mined by summing the unit costs of felling by chain-saw and the unit costs of winching, debranching andcross-cutting by processor.


Costs and efficiency of timber harvesting by NIAB 5–15 processor mounted on a farm tractor (177–184) J. SOWA et al.

Fig. 1 Timber debranching and cross-cutting

Slika 1. Kresanje grana i trupljenje

Fig. 2 Processed timber on the skid trail

Slika 2. Izra|eno drvo na traktorskoj vlaci

5. Results and discussion – Rezultati irasprava

On the basis of the above assumptions, the costsof exploitation of each machine used in the analysedtechnological process were calculated as follows:the chain saw: C = 2.28 �/h, the farm tractor (carrier):C = 6.75 �/h, the processor: C = 2.45 �/h, whichmeans that the costs of exploitation of the processorwith the tractor equal 9.20 �/h. A detailed cost struc-ture related to each machine is presented in Table 4.

The cost structure of the processor is dominatedby fixed costs because it is not fuelled but driven bytractor. The fixed costs constitute over 60% of thetotal cost. This percentage is less than 40% for thewhole assembly because the tractor, whose exploita-tion costs dominate in the assembly, has a share of29% in fixed costs. The conclusion is that by pro-longing the working period of the whole assembly,the costs of its exploitation will not be considerab-ly reduced. For example, work during 1.5 shifts(12 hours/day) will reduce the costs of exploitationby 15%, i.e. bring them down to 7.82 �/h.

Effectiveness on each work-stand, calculated onthe basis of the field time study, is presented inFigures 3 and 4.

In the analysed technology, the chainsaw opera-tor’s work consists exclusively of tree felling; that iswhy the effectiveness of his job was very high. In



Table 3 Assumed cost parameters for machine rate calculation

Tablica 3. Pretpostavljeni tro{kovni parametri za kalkulaciju strojnoga rada

Cost calculation inputs

Ulazni podaci kalkulacije

Chain saw

Motorna pila

NIAB 5–15 processor

Procesor NIAB 5–15

Farm tractor

Poljoprivredni traktor

Machine purchase price – Nabavna vrijednost stroja P [�] 658 21,053 26,316

Machine economic life T [years] – Normalno vrijeme uporabe stroja T [godine] 2 8 8

Scheduled operating time H [h/year] – Iskori{tenost stroja H [h/godina] 1800 1800 1800

Capital interest rate – Kamatna stopa p [%] 13.2 13.2 13.2

Fuel price – Cijena goriva [�/dm3] 1.11 – 1.03

Oil price – Cijena maziva [�/dm3] 1.8 1.8 –

Repair cost index – Indeks tro{kova popravka n [%] 30 30 10

Table 4 Cost structure per hour of machine

Tablica 4. Struktura tro{kova po satu rada stroja

Cost item

Sastavnice tro{ka

Chain saw

Motorna pila

NIAB 5–15 processor

NIAB 5–15 procesor

Farm tractor

Poljoprivredni traktor

Harvesting system

Sustav pridobivanja drva

Amortization – Amortizacija [�/h] 0.18 1.46 1.83 3.29

Cost of interest – Tro{ak kamata [�/h] 0.01 0.10 0.12 0.22

Fixed costs – Fiksni tro{kovi [�/h] 0.19 1.56 1.95 3.51

Fuel and lubrication – Gorivo i mazivo [�/h] 2.03 0.46 4.62 5.08

Repairs – Popravci [�/h] 0.05 0.44 0.18 0.62

Variable costs – Varijabilni tro{kovi [�/h] 2.08 0.90 4.80 5.70

Fig. 3 Productivity of the chainsaw operator in effective time

Slika 3. Proizvodnost radnika s motornom pilom u efektivnom vremenu

early thinnings, the effectiveness of felling was notvery differentiated and remained in the range from4.3 to 5.0 m3/h (Fig. 3). In late thinnings, the ef-fectiveness was higher than in the early ones, whichwas related to a greater volume of a single tree. Theeffectiveness of the processor was more differentiatedin early thinnings (Fig. 4). The highest effectiveness,i.e. almost 2 m3/h, was observed in fir forest while in

spruce and pine forests it was by about 0.7 m3/hlower. In late thinnings the effectiveness of the pro-cessor was very balanced and amounted to about3 m3/h. The results obtained are close to the onesobserved in Germany and Sweden, where, depend-ing on the stand composition and age, the effecti-veness was between 1.5 and 3.8 m3/h (Giefing 1994a,Marntell and Marntell 1988, Walczyk 1997).

Calculated on the basis of exploitation costs perhour and the effectiveness obtained, the unit costs oftree felling by chain saw are presented in Figure 5.

Low costs of the chain saw harvesting (Tab. 4)and its high effectiveness, connected with the use ofthis equipment for only one technological operation(i.e. felling), resulted in very low unit costs. In earlythinnings, they amounted to 0.48 �/m3 on averageand in late thinnings to 0.28 �/m3.

The work of the assembly with the processorconstitutes the chief item in the analysed technologycosts (Fig. 6).

The application of the processor in late thinningsresulted in the costs of 2.53 �/m3 in the operationaltime; and in early thinnings the costs were almosttwice higher: 4.74 �/m3. A larger differentiation ofunit costs, obtained in early thinnings carried out instands composed of different species, results from ahigher effectiveness of the processor in fir stands.

Summary unit costs in the operational time forthe whole technology are as follows: felling, winching,debranching and cross-cutting (Fig. 7) do not exceed6.44 �/m3 in younger stands (early thinnings) and



Fig. 4 Productivity of the processor in effective time

Slika 4. Proizvodnost procesora u efektivnom vremenu

Fig. 5 Unit costs of tree felling (chainsaw operator) in effective time

Slika 5. Jedini~ni tro{kovi sje~e stabala (radnik s motornom pilom) uefektivnom vremenu

Fig. 6 Unit costs of work of the processor NIAB 5–15 in effective time

Slika 6. Jedini~ni tro{kovi rada procesora NIAB 5–15 u efektivnomvremenu

2.95 �/m3 in older stands (late thinnings). The lack ofprocessors mounted on farm tractors in the Polishmarket of forest services (Kocel 2003) and, hence, thelack of economic analyses of such equipment makesit impossible to directly compare the results obtainedin the present research with those by other authors.Comparisons with the economic effectiveness of har-vesting of such equipment achieved abroad are notreliable due to considerable differences in the gene-ral economic situation, which affects the level of unitcosts. For example, according to Giefing (1994a) theunit costs of timber harvesting in Germany, in con-ditions very similar to stand No. 1 (Tab. 1), amount-ed to 23 DM/m3.

A simulation of total unit costs, obtained for thewhole shift, was performed for the analysed tech-nology. It was assumed that the monthly costs perperson for the chainsaw operator and the driver-operator (gross wages with surcharge) are 6.58 �. Inaccordance with the results, the coefficient of theshift utilisation was established on the level of 0.75.For such conditions, the average total unit costs are:10.67 �/m3 in early thinnings and 7.49 �/m3 in latethinnings.

The present research results concerning the ex-ploitation and effectiveness of the processor indicatethe existence of direct relations between them andthe average volume of harvested trees. These rela-tions are presented in Figure 8. The regression equa-tions shown there turned out to be statistically sig-

nificant. In the case of effectiveness, the significancelevel, calculated by means of the t-Student test, was0.00 (t = 4.11) and in the case of unit costs 0.00(t = –3.78). The strength of the relation between theaverage volume of harvested trees and effectivenessVo, measured by the value of the Pearson coefficientof linear correlation R, amounted to 0.72. In the caseof unit costs co, the correlation coefficient was –0.69.The straight lines of regression (Fig. 8) allow forpredicting the value of unit costs and effectivenessdepending on the average volume of harvested treesexclusively based on the range of the volume ofanalysed trees. This is due to two facts: firstly, theprocessor can debranch trees whose maximum dia-meter is 50 cm; secondly, using the processor toharvest trees with small mass, e.g. in late thinnings,would require additional research to be performedin such conditions because manual bunching of logswould have to be considered as well as a lot of workfor harvesting of timber of low market value.


On the basis of research results, the followingconclusions should be pointed out:

Þ Hourly costs of harvesting of the processorassembly amounted to about 9.2 �.

Þ Different values of effectiveness obtained inearly and late thinnings resulted in differentlevels of unit costs in these groups of stands.



Fig. 7 Unit costs of analysed technology (chainsaw operator + pro-cessor NIAB 5–15) in effective time

Slika 7. Jedini~ni tro{kovi istra`ivane tehnologije (radnik s motornompilom + procesor NIAB 5–15) u efektivnom vremenu

Fig. 8 Unit costs and productivity in effective time vs. the mean treevolume

Slika 8. Ovisnost jedini~noga tro{ka i proizvodnosti u efektivnom vre-menu o srednjem obujmu stabla

Þ The costs of harvesting of the tractor are thedominant item in the costs of harvesting of theprocessor-tractor assembly.

Þ Due to a small share of fixed costs in the costsof harvesting of the harvesting system, pro-longing the shift length from 8 to 12 h loweredthe unit costs only by 15%.

Þ Considering the relations between the aver-age volume of processed trees and effective-ness (and, hence, unit costs), as shown in thepresent research, it is more profitable to usethe processor in late thinnings.


Giefing, D. F., 1994a: Badania eksploatacyjne procesoraHYPRO. Przegl¹d Techniki Rolniczej i Leœnej 12: 20–22.

Giefing, D. F., 1994b: Ci¹gniki rolnicze w procesie pozyski-wania drewna. Przegl¹d Techniki Rolniczej i Leœnej 10:22–23.

Giefing, D. F., Gackowski, M., 2001: Ekonomiczna efek-tywnoœ} pozyskiwania drewna krótkiego w drzewosta-nach III kl. wieku w zale¿noœci od zastosowanych urz¹dzeñ

zrywkowych. Polska Akademia Umiejêtnoœci, Prace Ko-misji Nauk Rolniczych 3, p. 17–26.

Jod³owski, K., (2000): Technologia i technika pozyskiwa-nia drewna we wczesnych trzebie¿ach w drzewostanachsosnowych. In: Stan i perspektywy badañ z zakresu u¿yt-

kowania lasu. Instytut Badawczy Leœnictwa, Warszawa, p.97–108.

Kocel, J., 2003: Sektor us³ug leœnych – stopieñ i zakresprzemian, instrumenty wspierania oraz rola sektora w³agodzeniu bezrobocia. Mat. IBL, Warszawa.

Karlsson, T., 1988: System Skonsam – studie av ett småska-ligt mekaniserat gallringssystem. Sveriges lantbruksuni-versitet, Institutionen for skogstechnik 119, p. 1–100.

Maciak, A., Skar¿yñski, J., 1995: Efekty pracy ci¹gnikaURSUS 2812 przy zrywce drewna krótkiego. Przegl¹d Tech-niki Rolniczej i Leœnej 10, p. 27–29.

Marntell, A., Marntell, T., 1988: Vimek G30 i praktisk drift.Vimek G30 in operation. Sveriges lantbruksuniversitet,Institutionen for skogstechnik, p. 1–45.

Porter, B., 1998: Analiza wybranych metod pozyskiwaniadrewna w ciêciach jednostkowych w drzewostanach œwier-kowych. Przegl¹d Techniki Rolniczej i Leœnej 5: 10–12.

Suwa³a, M., 1998: Koszty pracy wybranych œrodków dopozyskiwania drewna. Sylwan 142(11): 27–36.

Suwa³a, M., Rzadkowski, S., 2001: Wydajnoœ} pracy, kosztyi uszkodzenia drzew przy pozyskiwaniu drewna w trze-bie¿ach drzewostanów górskich. Prace IBL, Seria A, 1 (911),p. 85–111.

Walczyk, J., 1997: Procesory tworzone na bazie ci¹gnikówrolniczych. Przegl¹d Techniki Rolniczej i Leœnej 3: 18–21.

Zychowicz, W., 1998: Metoda obliczania kosztów eksploa-tacji maszyn leœnych. In: Mat. Symp. Efekty stosowaniamaszyn o du¿ej wydajnoœci, przyjaznych dla œrodowiskaw lasach polskich, Warszawa, p. 5–12.

Saètak

Djelotvornost pridobivanja drva procesorom NIAB 5–15 postavljenimna poljoprivrednom traktoru

Rad prikazuje istraìvanje djelotvornosti pridobivanja drva procesorom NIAB 5–15 postavljenim na poljopri-vrednom traktoru. U Poljskoj postoji velik broj poduze}a koja se bave {umskim radovima. Ve}inom su to mala po-duze}a ograni~enih financijskih mogu}nosti te s nedostatkom visokou~inkovitih, ali vrlo skupih {umskih strojeva.Pove}anje stupnja mehaniziranosti sje~e i izrade stabala omogu}eno je primjenom procesora postavljenoga napoljoprivrednom traktoru. Na slobodnom trì{tu jedan je od osnovnih kriterija primjene odre|enoga postupkapridobivanja drva ekonomska isplativost. Stoga je potrebno detaljno analizirati tro{kove sustava rada, {to je ujednoi cilj ovoga istraìvanja.

Istraìvanje je ograni~eno na rane i kasne prorede borovih, jelovih i smrekovih sastojina. Mjerenja su provedenana pokusnim plohama veli~ine 0,5 ha uz traktorsku vlaku. Osnovne zna~ajke istraìvanih sastojina prikazane su utablici 1, a tehni~ki podaci o ispitivanom procesoru u tablici 2.

Postupak pridobivanja drva podrazumijeva usmjerenu sje~u dozna~enih stabala motornom pilom okomitoprema smjeru {umske vlake, vezanje tovara i privitlavanje vitlom, kresanje grana i trupljenje debla procesorom.Drvni su se sortimenti izra|ivali u duljinama od 1,25 m.

Studij rada i vremena proveden je proto~nom metodom zasebno za radnika s motornom pilom i radnika naprocesoru. Izra|enim drvnim sortimentima odre|en je obujam na osnovi mjerenja srednjega promjera. Tako|er jemjerena potro{nja goriva i maziva motorne pile i traktora.



Tro{kovi su izra~unati na osnovi efektivnoga vremena rada. Za potrebe istraìvanja pretpostavljeno je da sustrojevi novi te da su u izra~unu uzete vrijednosti prikazane u tablici 3.

Satni tro{ak pojedinoga stroja pri ispitivanom postupku iznosi: motorna pila 2,28 EUR/h, poljoprivrednitraktor 6,75 EUR/h, procesor 2,45 EUR/h, {to daje ukupni tro{ak od 9,20 EUR/h. Detaljna je struktura tro{kovaprikazana u tablici 4. U strukturi tro{kova procesora prevladavaju fiksni tro{kovi (preko 60 %), jer procesor ne tro{igorivo ve} ga pogoni traktor. Za cijeli sustav pridobivanja drva fiksni tro{kovi iznose manje od 40 % jer traktor, ~ijisu tro{kovi najve}i, ima samo 29 % fiksnih tro{kova. Zaklju~ak je da produljenje radnoga vremena ne}e bitnoutjecati na smanjenje tro{kova (npr. radnim danom od 12 sati mogu se smanjiti tro{kovi za samo 15 %).

Proizvodnost radnika s motornom pilom i radnika na procesoru izra~unata je na osnovi provedenoga studijarada i vremena. Radnik s motornom pilom radi isklju~ivo na sje~i stabala te je stoga njegova proizvodnost velika ikre}e se od 4,3 m3/h do 5 m3/h u ranim proredama. U kasnim proredama proizvodnost radnika s motornom pilomjo{ je ve}a zbog ve}ega obujma stabala (slika 3). Proizvodnost radnika na procesoru razlikuje se kod ranih proredaovisno o vrsti drve}a, a u kasnim proredama iznosi pribli`no 3 m3/h u uvjetima rada obuhva}enim istraìvanjem(slika 4).

Jedini~ni su tro{kovi izra~unati na osnovi satnoga tro{ka rada pojedinoga sredstva i ostvarene proizvodnosti.Rad motornom pilom ostvario je male jedini~ne tro{kove (0,48 EUR/m3 u ranim proredama, odnosno 0,28 EUR/m3

u kasnim proredama) zbog niskoga satnoga tro{ka i velike prizvodnosti (slika 5). Najve}i su jedini~ni tro{kovizabiljeèni u radu na procesoru – 4,74 EUR/m3 u ranim proredama i 2,53 EUR/m3 u kasnim proredama (slika 6).Ukupni jedini~ni tro{kovi sustava pridobivanja drva ne iznose vi{e od 6,44 EUR/m3 u ranim proredama, odnosno2,95 EUR/m3 u kasnim proredama (slika 7).

Provedeno je istraìvanje pokazalo ovisnost proizvodnosti procesora te jedini~nih tro{kova o obujmu posje~eno-ga stabla. Stoga je provedena regresijska analiza navedenih parova podataka te se stati~kim ispitivanjem utvrdilazna~ajna povezanost podataka. Rezultati regresijske analize omogu}uju predvi|anje vrijednosti jedini~nih tro{ko-va i proizvodnosti rada procesora ovisno o srednjem obujmu posje~enoga stabla. Na osnovi promatranih odnosazaklju~uje se da je ve}a isplativosti rada procesora u kasnim proredama.

Klju~ne rije~i: pridobivanje drva, prorede, procesor, tro{kovi, proizvodnost



Authors’ address – Adresa autorâ:

Prof. Janusz Michal Sowa, PhD.e-mail: [email protected] Kulak, PhD.e-mail: [email protected] Szewczyk, MSc.e-mail: [email protected] University of CracowFaculty of ForestryDepartment of Forest and Wood UtilizationAl. 29 Listopada 4631–425 KrakówPOLAND

Received (Primljeno): December 19, 2006Accepted (Prihva}eno): November 19, 2007

Optimization of an existing forest roadnetwork using Network 2000

Mohammad Reza Ghaffarian, Hooshang Sobhani


Optimization of forest road network is an important part of logging planning. Matthews (1942)was the first to introduce a method for optimization of road spacing based on minimizationof roading and skidding cost. The goal of this paper is to find the best road network for adistrict harvested by skidder. The skidding model developed by stepwise regression modelwas used to predict the cost of skidding per cubic meter for the 39 nodes, which were plannedin the district map. The harvesting volume and roading cost per each node were computed.The data were entered into NETWORK 2000 and shortest path algorithm; simulatedannealing and great deluge algorithms were run to find the best solution to optimize loggingcost of the district. The results showed which roads can be eliminated from the existing forestroad network.

Keywords: road spacing, optimal road density, skidder, model of network analysis, Iran


Forest road network planning is an importanttask of forest engineers. Matthews (1942) was thefirst to present a model for defining optimum roadspacing based on minimization of skidding and road-ing costs from a landowner’s point of view. Optimalroad spacing can also be influenced by other factorshaving impact on optimal road network such aslogging method, price of products, taxation policies,landing costs, overhead costs, equipment opportu-nity costs, road width and size of landing, skiddingpattern, profit of logging contractor, slope and topo-graphy and soil disturbance (Segebaden 1964, Sund-berg 1976, Peters 1978, Bryer 1983, Wenger 1984, Ses-sions 1986, Yeap and Sessions 1988, Thompson 1992,Liu and Corcoran 1993, Heinimann 1997, Thomp-son 1998, Akay and Sessions 2001, Sessions andBoston 2006).

A couple of studies on optimal road density (ORD)have been reported by several researchers. Pi~manand Pentek (1998) calculated ORD of 14.7 m/ha forground based skidding system using farm tractorsin Croatia. In the Northern forests of Iran, the casestudies on selection cuttings and skidding opera-tions showed that optimal road density ranged from9 to 28 m/ha for different areas (Mostafanejad 1995,Eghtesadi 2000, Lotfalian 2001, Naghdi 2004). Most

of the above studies used minimization of total costof roading and skidding. Optimal road spacing isonly a value that provides a guide for locating roadsand cost target but does not suggest where the roadsshould be actually placed (Tan 1999). In the pastyears, the mixed integer mathematical programmingand heuristic algorithms such as: TIMBRI (Sullivan1974), NETWORK (Sessions 1978), TRANSHIP (Kir-bey et al. 1981), MINCOST (Wong 1981), NETCOST(Weintraub 1990) and NETWORK 2000 (Sessions andChung 2003) have been used to find the appropriatesolution for certain fixed and variable cost problem.Sessions (1992) introduced the method of using net-work analysis for road and harvesting planning, whichis applied in this study. Gullison and Hardner (1993)described a rule-based simulation model designedto examine options for reducing the total length offorest roads. Clarck et al. 2000) also used a heuristicfor access road development where roads are de-fined a priori. Akay et al. (2005) described com-monly used modern heuristic techniques (SimulatedAnnealing, Genetic Algorithm, Tabu Search, and Short-est Path Algorithm). Simulated Annealing (SA) wasused to guide the search for the best vertical align-ment that minimizes the total costs of construction,transportation, and maintenance costs for a singleforest road. Ichihara et al. (1996) proposed GeneticAlgorithm (GA) model integrating two optimization


Preliminary note – Prethodno priop}enje

techniques to optimize forest road profiles. TabuSearch and GA were compared to manually design-ed forest road profile by Aruga et al. (2005). Short-est path algorithms have been implemented bySessions (1991) for secondary harvest transport. An-derson and Nelson (2004) used Dijkrstra’s shortestpath algorithm to project a road link that minimizedthe distance between a landing and the current roadnetwork.

Tan (1999) developed the spatial and heuristicprocedure to locate forest roads. He reported that theimproved procedure proved to be beneficial in help-ing forest road planning managers evaluate alter-natives and hence select the optimal location for aroad network.

Stueckelberger et al. (2006) considered roadingcost, ecological effects and suitability for cable yard-ing landings in their automatic road-network planningusing multi-objective optimization in Switzerland.

The current paper evaluates an existing forestroad network in a district of research forest in Nort-hern Iran using Network analysis. The goal is to findthe best road network which minimizes total cost ofskidding and roading in this district. The results ofnetwork procedure can show which road segmentscan be eliminated.

2. Method of study – Metode istra`ivanja

2.1 Site of study – Mjesto istra`ivanja

The study site is located in Northern forests ofIran in Nowshahr. The research was carried out onroad network of Namkhaneh research district andtraining forest centre of Tehran University. The ma-nagement method is mixed un-even aged high forestwith single and group selective cutting regime. Thedistrict covers the area of 1,080.9 ha with the grow-ing stock of 434 m3/ha. The broadleaves stands most-ly consist of Fagus sp., Quercus sp., Carpinus sp. Thecutting volume of the district is 5,850 m3 per yearwhich means 5.41 m3/ha. The slope ranges from 15to 60%. The felling is done motor-manually usingchainsaws. The felled trees are then delimbed andbucked to the assortments. The sawmill logs areskidded by wheeled cable skidders or tracked skid-ders to the roadside landings. The fuel woods areextracted by mules. Also, in steep terrain that cannot be harvested by skidders, logs are processed tosmall lumber so as to be extracted by mules. Skid-ding group includes operator and chokerman.

2.2 Data collection – Prikupljanje podataka

Jour Gholami (2005) studied cost production ofTAF 1004P and Timberjack 450C skidders in this

area. He used the continuous time study method inboth production studies. A typical work cycle in-cluded: travel empty, releasing the winch, chokersetting, winching, travel loaded, unhooking and dack-ing. During this time study technical, personal andoperational delays were recorded. The same variableswere used in both data collections. Variables includ-ed skidding distance, piece volume, load volume,number of pieces per turn and slope of trail. In thistime study forty-four cycles for Timberjack 450C andforty-six for TAF 1004P were collected. Using thestepwise regression method, the time predicting mo-del was developed for each cycle.

The system cost of the skidder was about 46.91 �/h(Jour Gholami 2005). The developed regression mo-dels to predict the time of skidding were used tostudy the road spacing.

The roading cost in this forestry centre included thecosts of planning (384.6 �/km), construction of subbase(6,837.6 �/km), subgrade construction (11,794.8 �/km),culverts and ditches (4,273.5 �/km), and maintenanceand repair cost as 15% of the roading cost in the10 years period (349.3 �/km). The road the con-struction cost was 23,639.8 �/km or 23.64 Euro/m.The interest of investment would be 2.19 �/m consi-dering 18.5% interest rate in Iran. If the life time ofthe forest road were 50 years, the deprecation costwould be 0.47 �/m. The total annual cost of roadingincludes the sum of the interest and deprecationcosts of 2.66 �/m.

The mean harvesting volume of Namkhaneh dis-trict is about 5.41 m3/ha/year. The existing road net-work has a density of 28.16 m/ha. The compartments201, 202, 203, 204 and 205 are protected because ofsteep slopes (Fig. 1).

Network 2000 was developed to optimize largefixed and variable cost problems related to transpor-tation. It provides three different algorithms: shortpath algorithm, simulated annealing (SA) and greatdeluge (GD). The first algorithm solves the networkproblem using a heuristic method that prorates thefixed costs in an iterative mode. The algorithm cansolve very quickly a large fixed and variable costproblem related to transportation. However, the bestsolution found by this algorithm might not be opti-mal, but it is a »good« solution. If the solution is notoptimal, it is possible to search for a better solutionwith continues iterations in this program.

To solve the problems of shortest path length, aDutch computer scientist named Dijkstra, develop-ed the shortest path algorithm in 1959. The basicpremise of this algorithm is to find the length of theshortest path between the starting vertex and the firstvertex; then the length of the shortest path betweenthe staring vertex and the second vertex; continuing


M. R. GHAFFARIAN and H. SOBHANI Optimization of an existing forest road network using Network 2000 (185–193)

until the length of the shortest path between thestarting vertex and ending vertex is found.

Simulated annealing (SA) is a search techniquewhich exploits an analogy between the way in whicha metal cools and freezes into a minimum energycrystalline structure (the annealing process) and thesearch for a minimum in a more general system; itforms the basis of an optimization technique forcombinatorial problems, etc. It was developed in1983 to deal with highly nonlinear problems. SA’smajor advantage over other methods is an ability toavoid becoming trapped in local minima. The algo-rithm employs a random search, which not only ac-cepts changes that decrease the objective function f(assuming a minimization problem), but also somechanges that increase it.

The great deluge algorithm (GDA) is a recentlydeveloped variant of simulated annealing. It is simi-lar to SA in that only a single change is considered asa »current« solution, the resulting temporary solu-tion is evaluated, and a decision is made whether ornot to convert the temporary solution to the currentsolution (Bettinger et al. 2002).

The GDA was introduced by Dueck (1993) andproved superior to similar Monte-Carlo based algo-rithms in solving a 442-city and 532-city TravelingSalesman Problem. The form of the GDA as pre-sented by Dueck (1993) consisted of using a singleparameter in the determining of whether or not to

convert the temporary solution to the current solu-tion (and perhaps change to an inferior solution).The use of one parameter rather than two, as in asimulated annealing algorithm, is believed to de-sensitize the algorithm thus leading to equally goodresults even when parameter estimation and formu-lation is poor.

In order to use Network 2000 program, the mapof compartments was used including the existingforest road in the district (Fig. 1). Each compartmentwas used to define the nodes for the network ap-proach. It was assumed that 2/3 of the logs in thecompartments are skidded downhill and the rest areskidded uphill. Based on this assumption, the nodeswere determined using Arc Map (Fig. 2). The area ofthe nodes was calculated, and then the logging volu-me for the nodes was determined based on its areaand logging volume of the management area. Thelogs are transported from all the nodes to 3 mills whichare marked in Fig. 2.

In each node the mean skidding distance was com-puted based on measuring 10 samples on the map. Theskidding cost per cubic meter (Variable cost) was com-puted for each node using the mean skidding distance,hourly cost and time predicting model.

The road length for each node was measured onthe map. The roading cost per node was computedby multiplying the road length and the roading costper unit length. In Network 2000 program, the Link


Optimization of an existing forest road network using Network 2000 (185–193) M. R. GHAFFARIAN and H. SOBHANI

Fig. 1 Existing road network of Namkhaneh district

Slika1. Postoje}a mre`a {umskih cesta podru~ja Namkhaneh

file provides the variable cost and upper and lowervolume bounds from node to node.

The Sale file includes entry nodes, destinationnode, harvesting volume, harvest year and discountrate.

The objective is to minimize skidding and roadingcost. The network model can be expressed mathe-matically as:

Minimize z = Sc V Rc Vi i

i

i

i ii

i

�

�

�

�

�

� �0

39

0

39

Subject to: Vi > 0

where:

Sci – Skidding cost per node per m3

Rci – Roading cost per segment

Vi – Harvesting volume per node

0 < i < 120.

3. Results – Rezultati

3.1 Production – Proizvodnost

The time study resulted in a net production rateof 8.22 m3/h (based on free delay hours). The grossproduction was 8.88 m3/h (including delay times).The skidding cost based on the net production ratewas 5.69 �/m3 (Jour Gholami 2005).

3.2 Time predicting model for skidding – Modelprocjene vremena privla~enja

Jour Gholami (2005) used stepwise regression met-hod using variables such as skidding distance, piece

volume, load volume, number of pieces per turn andslope of the skid trail. The model was developed bySPSS.

t = 4.142 + 1.988 × N + 0.0176 × L +

+ 1.093 × V (R2 = 0.786, n = 43)

where:

t – skidding time, min/cycle

N – number of pieces per cycle

L – skidding distance, m

V – volume of load, m3/cycle

The effect of each variable on skidding time wasstudied by changing one variable while holding theother variables constant at their mean value. Figures3, 4 and 5 show the effect of skidding distance, num-ber of pieces per turn and load volume on skiddingtime, respectively.

Increasing skidding distance, number of pieces perturn and load volume will increase skidding time.

3.3 Optimum road network – Optimalna mre`aprometnica

The variable cost (skidding cost per m3) and fixedcost (road construction cost per segment) were en-tered for all the nodes in the Link file. For the Salefile, the entry node, destination node and harvestingvolume per node were entered. The same harvestyear was assumed for all nodes.

Shortest path algorithm was run based on thedata and the best solution was found at 45 iteration



Fig. 2 Nodes and mills of network analysis

Slika 2. ^vori{ta i pilane u mre`noj analizi

of the shortest path algorithm. Figure 6 shows thelinks to be used based on the result of running thealgorithm. Using these results, the proposed linkswere marked on the map and the best solution wasobtained as shown in Figure 7.

Total variable cost, fixed cost and sum of vari-able and fixed cost was 8.30 �/m3, 18.89 �/m3 and

27.19 �/m3, respectively. SA and GDA were also run

to see if they can find better solution, but these al-gorithms could not find better solution than theshortest path algorithm.

The network analysis suggested eliminating theroad from node 1 to 2 and the road from compart-ment 28 to 38.

4. Discussion – Rasprava

The results of work and time study indicated thatthe variables such as number of pieces per turn, loadvolume and skidding distance have significant effecton RMS of the model. If these variables increase, thecycle time and cost of skidding will increase.

The solution found by running the shortest pathalgorithm helps logging planners decide which roadsegments can be used and which segments should beeliminated to achieve the minimum total cost ofroading and skidding.

The solution suggested eliminating the road seg-ments of nodes 1, 2, 28 and 38. If these roads areclosed, the forest company not only can avoid theirmaintenance cost but can also decrease total cost ofskidding and roading of the district. The logs innodes 1 and 2 can be extracted to nodes 0 and 3using the longer skidding distance. The felled treesin nodes 28, 29 and 38 should be skidded to node 16and 27.



Fig. 4 Cycle time vs. number of pieces per turn

Slika 4. Ovisnost vremena turnusa o broju komada u tovaru

Fig. 5 Cycle time vs. load volume

Slika 5. Ovisnost vremena turnusa o obujmu tovara

Fig. 3 Cycle time vs. skidding distance

Slika 3. Ovisnost vremena turnusa o udaljenosti privla~enja



Fig. 6 Links in network analysis

Slika 6. Poveznice u mre`noj analizi

Fig. 7 Best solution of Network Analysis for the road network of Namkhaneh district

Slika 7. Najpovoljnije rje{enje mre`e {umskih cesta u podru~ju Namkhaneh dobiveno mre`nom analizom


The results of time study and the model deve-loped for predicting skidding time give the forestengineers a tool for logging planning.

Network analysis using the shortest path algo-rithm, GD or SA, is a useful method to optimizetransportation problems. The results of running Net-work 2000 based on an existing forest road networkharvested by cable skidder, showed that the bestsolution (Fig. 7) can be found with the total cost of27.19 �/m3. The solution showed which links can beused to achieve the lowest cost of skidding androading (Fig. 6). Using the proposed links, it wasestablished which road segments can be eliminated.

Acknowledgement – Zahvala

The authors wish to thank Prof. John Sessionsfrom Oregon State University for his help in apply-ing Network Analysis in this paper. The authorswould also like to thank Mr. Martin Kuehmaier forhis support.


Akay, A., Boston, K., Sessions, J., 2005: The evolution ofcomputer-aided road design systems. International Jour-nal of Forest Engineering 16(2): 73–79.

Akay, A., Sessions, J., 2001: Minimizing road constructionplus forwarding costs under a maximum soil disturbanceconstraint. The International Mountain Logging and 11th

Pacific Northwest Skyline Symposium December 10–12,Seattle, Washington, USA, p. 268–279.

Anderson, A. E., Sessions, J., 2004: Projecting vector basedroad networks with a shortest path algorithm. Can. J. For.Res. 34(7): 1444–1457.

Aruga, K., Sessions, J., Akay, A., 2005: Heuristics techni-ques applied to forest road profile. Japanese Forestry So-ciety, Journal of Forest Research 10(2): 83–92.

Bettinger, P., Graetz, D., Boston, K., Sessions, J., Chung, W.,2002: Eight heuristic planning techniques applied to threeincreasingly difficult wildlife planning problems. Silva Fen-nica 36(2): 561–584.

Bryer, J. B., 1983: The effects of a Geometric Redefinition ofthe Classical Road and Landing Spacing Model ThroughShifting. Journal of Forest Science 29(3): 670-674.

Clark, M. M., Meller, R. D., McDonald, T. P., 2000: A three-stage heuristic for harvest scheduling with access roadnetwork development. Forest Science 46: 204–218.

Dueck, G., 1993: New optimization heuristics: The greatdeluge algorithm and the record-to-record travel. Journalof Computational Physics 104: 86–92.

Gullison, R. E., Hardner, J. J., 1993: The effect of road designand harvest intensity on forest damage caused by selectivelogging: Empirical results and a simulation model fromthe Bosque Chimanes, Bolivia. Forest Ecology and Mange-ment 59: 1–14.

Eghtesadi, A., 2000: Study of transportation network andmachinery in Vaz Forest Area. PhD. Thesis, Azad University.

Heinimann, H. R., 1997: A computer model to differentiateskidder and cable-yarder based road network concepts onsteep slopes. Journal of Forest Research (Japan) 3(1): 1–9.

Ichihara, K., Tanaka, T., Sawaguchi, I., Umeda, S., Toyoka-wa, K., 1996: The method for the profile of forest roadssupported by genetic algorithm. The Japanese ForestrySociety, Journal of Forest Research 1: 45–49.

Jour Gholami, M., 2005: Study of efficiency, productionand cost of the large and small skidders (Case study of Tafand Timberjack 450C). MSc. Thesis, Faculty of NaturalResources, Tehran University.

Kirby, M., Wong, P., Hager, W., 1981: Guide to Transhipmodel. USDA For. Serv. Pac. Southwest For. Range Exp.Stn. Berkeley.

Liu, S., Corcoran, T. J., 1993: Road and landing spacingunder the consideration of surface dimension of road andlandings. Journal of Forest Engineering 5(1): 49–53.

Lotfalian, M. 2001: Study of the factors influencing opti-mum road density in Sangdeh-Mazandaran. PhD. Thesis,Tehran University, Faculty of Natural Resources.

Matthews, D. M., 1942: Cost control in the logging industry.McGraw-Hill, New York, 374 p.

Mostafanejad, A., 1995: Study of cost production of skidderTimberjack 450C and optimal length of skid trails. MasterThesis, Tehran University, Faculty of Natural Resources.

Naghdi, R., 2004: Study of optimum road density in treelength and cut to length system, PhD. Thesis, University ofTarbiat Modarres, Faculty of Natural Resources.

Pi~man, D., Pentek, T., 1998: The influence of forest roadsbuilding and maintenance costs on their optimum densityin low lying forests of Croatia. Proceedings of the Seminaron Environmentally Sound Forest Roads and Wood Trans-port in Sinaia, Romania, FAO Rome, 87–102.

Segebaden, G. V., 1964: Studies of cross-country transpor-tation distances and road net extension. Studia ForestaliaSuecica, Nr. 18

Sessions, J., 1986: Can income tax rules affect managementstrategies for forest roads. Western Journal of Applied Fo-restry 1(1): 26–28.

Sessions, J., 1978: A Heuristic Algorithm for the Solution ofthe Variable and Fixed Cost Transportation Problem. Inproceeding: The 1985 Symposium on Microcomputers.Journal of Forest Engineering 10(2): 91–100.

Thompson, M. A., 1992: Considering overhead costs in roadand landing spacing models. Journal of Forest Engineer-ing 3(2): 13–19.



Thompson, M. A., 1988: Optimizing spur road spacing onthe basis of profit potential. Forest Product Journal 38(5):53–57.

Peters, P. A., 1978: Spacing of roads and landings to mini-mize timber harvest cost. Journal of Forest Science 24(2):209–217.

Yeap, Y. H., Sessions, J., 1988: Optimizing spacing andstandards of logging roads on uniform terrain. Journal ofTropical Forest Sceince 1(3): 215–228.

Weintraub, A., 1986: NETCOST, a heuristic approach forroading and forest management planning. COOP Agree-ment Rep., Berkeley, Univ. California.

Wenger, K., 1984: Cost control formulas for logging opera-tions. p. 560–563 in Forestry Handbook, 2nd Edition. So-ciety of American Foresters. John Wiley and Sons. NewYork. 1335 p.

Wong, P., 1981: An empirical evaluation of the prorationoption of the MINCOST network program. USDA ForestService Engineering Field Notes. Washington D.C. pp.15–22.

Saètak

Optimizacija postoje}e mreè {umskih cesta aplikacijom Network 2000

Planiranje mreè {umskih cesta, odnosno njihova optimizacija jedan je od vrlo va`nih zadataka inènjera {u-marstva. Mathews je (1942) prvi autor koji je predstavio model optimizacije mreè {umskih cesta (optimalnogarazmaka izme|u {umskih cesta) zasnovan na najmanjim ukupnim tro{kovima (s jedne strane tro{kova privla~enjadrva, a s druge strane svih tro{kova povezanih s uspostavom {umskih cesta na terenu).

Mnogi su autori (Sundberg 1976, Peters 1978, Bryer 1983, Wenger 1984, Sessions 1986, Yeap i Sessions 1988,Thompson 1992, Liu i Corcoran 1993, Segebaden 1964, Heinimann 1997, Thompson 1998, Akay i Sessions 2001,Sessions i Boston 2006) odredili osnovne ~imbenike koji utje~u na optimalan razmak {umskih cesta. To su: metodasje~e i izradbe, tehnologija pridobivanja drva, cijena drvnih proizvoda, politika oporezivanja, tro{kovi skladi{tenjatrupaca, op}i tro{kovi, oblik sekundarne mreè {umskih prometnica, dobit izvoditelja radova pridobivanja drva, {i-rina tijela {umskih cesta, veli~ina stovari{ta, nagib terena, konfiguracija terena i mogu}a o{te}enja tla.

U prebornim je {umama sjevernoga Irana, u razli~itim reljefnim podru~jima, izra|eno vi{e studija (Mostafane-jad 1995, Eghtesadi 2000, Lotfalian 2001, Naghdi 2004) kojima je utvr|ena optimalna gusto}a mreè {umskih cestau rasponu od 9 do 28 m/ha. Pri tome je primijenjena metoda najmanjih ukupnih tro{kova pridobivanja drva.

Optimalan razmak izme|u {umskih cesta, s jedne strane, predstavlja broj~anu vrijednost koja je putokaz i vo-dilja {umarskim stru~njacima pri provedbi primarnoga otvaranja {uma, ali s druge strane ne kazuje ni{ta o stvar-nim trasama {umskih cesta na terenu (Tan 1999). Radi optimizacije mreè {umskih cesta i odre|ivanja optimal-noga poloàja svake pojedine trase {umske ceste razvijen je ve}i broj matemati~kih programa i algoritama: TIMBRI– Sullivan 1974, TRANSHIP – Kirbey i dr. 1981, MINCOST – Wong 1981, NETCOST – Weintraub 1990,NETWORK – Sessions 1978 te NETWORK 2000 – Sessions i Chung 2003.

U ovom je radu, za mre`nu analizu postoje}ega primarnoga sustava {umskih cesta, primijenjen ra~unalni pro-gramski paket NETWORK 2000 koji je dizajniran radi optimizacije fiksnih i varijabilnih tro{kova koji se javljajupri transportu drva, a sastoji se od triju razli~itih algoritama. Definiran je i cilj istraìvanja – prona}i, u okviru po-stoje}e primarne {umske prometne infrastrukture, najbolju mreù {umskih cesta sa stajali{ta najmanjih ukupnihtro{kova pridobivanja drva (tro{kova privla~enja drva i tro{kova povezanih sa {umskim cestama).

Podru~je istraìvanja – gospodarska jedinica Namkhaneh (1080,90 ha) smje{tena je u mje{ovitim prebornim{umama sjevernoga Irana. To je nastavno-pokusni {umski objekt Sveu~ili{ta u Teheranu. Prosje~na drvna zalihaiznosi 434 m3/ha, a godi{nji je etat 5,41 m3/ha (oko 5850 m3 na ~itavom istraìvanom podru~ju). Sje~a i izradbaobavlja su motornim pilama lan~anicama uz primjenu. Pilanska se oblovina privla~i do pomo}nih stovari{ta kota-~nim i gusjeni~nim skiderima. Jednometarsko ogrjevno drvo iznosi se mulama kao i sve izra|eno drvo na strmimterenima koji su nedostupni skiderima (tada se od pilanske oblovine na mjestu sje~e izra|uju piljenice). Nagibi seterena kre}u od 15 do 60 %. Postoje}a primarna otvorenost {uma iznosi 28,16 m/ha uz napomenu da su odjeli 201,202, 203, 204 i 205 za{titni zbog velikoga nagiba terena te nisu obuhva}eni daljnjim analizama.

U ovom istraìvanju kori{teni su rezultati prija{njih istraìvanja. Jour Gholami (2005) do{ao je do ovih rezul-tata: u~inak zglobnoga traktora Timberjack 450C iznosi 8,22 m3/h, tro{ak strojnoga rada 46,91 EUR/h, a jedini~nitro{ak 5,69 EUR/m3. Isti je autor razvio i regresijski model za odre|ivanje utro{ka vremena pojedinoga turnusaprivla~enja drva, ~iji su ulazni parametri: udaljenost privla~enja, nagib traktorskoga puta (vlake), obujam pojedi-



noga trupca, obujam tovara i broj komada trupaca u tovaru. Utjecaj udaljenosti privla~enja, broja komada trupacau tovaru i obujma tovara na trajanje turnusa privla~enja drva prikazan je na slikama 3, 4 i 5. Smanjenje srednjeudaljenosti privla~enja, manji broj komada trupaca u tovaru i manji obujam tovara utje~u na smanjenje turnusaprivla~enja drva.

Tro{kovi povezani sa {umskim cestama uklju~uju: projektiranje (384,6 EUR/km), izgradnju donjega stroja(6837,6 EUR/km), izvedbu gornjega stroja (11 794,8 EUR/km), izradu objekata odvodnje – cijevnih propusta i od-vodnih jaraka (4273,5 EUR/km) i odràvanje (349,3 EUR/km). Ukupni je tro{ak povezan sa {umskim cestama i iz-nosi 23 639,8 EUR/km. Uz razdoblje amortizacije od 50 godina i vaè}e financijske pokazatelje u Iranu godi{nji jetro{ak povezan sa {umskim cestama 2,66 EUR/m.

Svaki odjel istraìvane gospodarske jedinice razdijeljen je, pomo}u programa ArcMap, na odre|eni broj gravi-tacijskih podru~ja, uz pretpostavku privla~enja 2/3 etata smjerom nizbrdo te 1/3 smjerom uzbrdo. Teì{te pojedino-ga gravitacijskoga podru~ja predstavljalo je ~vor (nod) mreè, poslije kori{ten pri mre`noj analizi.

Sva je izra|ena pilanska oblovina transportirana do jedne od triju pilana ~iji je poloàj ozna~en na slici 2. Sred-nja je udaljenost privla~enja odre|ena izmjerom na digitalnom zemljovidu. Na zemljovidu je odre|ena i duljina{umskih cesta pojedinoga ~vora. Prema navedenim modelima i cijenama izra~unati su, primjenom matemati~kogaoblika mre`noga modela, tro{kovi privla~enja drva i tro{kovi povezani sa {umskim cestama. Cilj je minimaliziratiukupne tro{kove pridobivanja drva.

Unosom i obradom podataka za svaki ~vor te pokretanjem algoritma za traènje najkra}ega puta (shortest pathalgorithm) dobivena je najbolja povezanost prikazana na slici 6. Kori{tenjem navedenoga rje{enja preporu~ena jeoptimalna mreà primarnih {umskih prometnica (slika 7).

Ukupni su varijabilni tro{kovi (tro{kovi privla~enja drva) 8,30 EUR/m3, varijabilni tro{kovi (tro{kovi poveza-ni sa {umskim cestama) 18,89 EUR/m3, odnosno ukupni tro{kovi pridobivanja drva 27,19 EUR/m3. Radi provjeredobivenih rezultata pokrenuti su i algoritmi SA (simulated annealing algorithm) i GDA (great deluge algorithm) uokviru programskoga paketa NETWORK 2000, ali oni nisu prona{li bolje rje{enje za primarno otvaranje {uma odonoga koje je ponudio algoritam za traènje najkra}ega puta.

Mre`nom analizom sugerirana je eliminacija {umskih cesta od ~vora 1 do ~vora 2 te od ~vora 28 do ~vora 38.Privla~enje izra|enih trupaca iz ~vora 1 i ~vora 2 treba obavljati prema pomo}nim stovari{tima u ~vorovima 0 i 3,dok trupce iz ~vorova 28 i 29 treba privla~iti prema stovari{tima u ~vorovima 16 i 27. Na taj su na~in ukupnitro{kovi pridobivanja drva za istraìvano podru~je minimalizirani.

Klju~ne rije~i: raspored {umskih cesta, optimalna gusto}a {umskih prometnica, skider, model mre`neanalize, Iran




Mohammad Reza Ghaffarian, MSc.e-mail: [email protected] of Natural Resources and Applied

Life Sciences ViennaDepartment of Forest and Soil SciencesInstitute of Forest EngineeringPeter Jordan Strasse 821190 ViennaAUSTRIA

Assoc. Prof. Hooshang Sobhani, PhD.e-mail: [email protected] UniversityCollege of Natural ResourcesDepartment of ForestryP.O. Box 31585–4314KarajIRAN

Received (Primljeno): October 16, 2007Accepted (Prihva}eno): December 6, 2007

ISSN 1845-5719

� ��

................................................................... 137

........................................... 145

............................................................................ 157

............................................... 169

................................................................. 177

........................ 185

2

Original scientific papers –

Preliminary note –

Izvorni znanstveni radovi

Prethodno priop}enje

STANIMIR STOILOVImprovement of wheel skidder tractive performance by tire inflation

pressure and tire chains

TOBIAS CREMER, BORJA VELAZQUEZ-MARTIEvaluation of two harvesting systems for the supply of wood-chips

in Norway spruce forests affected by bark beetles

ANTON POJE, IGOR POTO^NIKInfluence of working conditions on overlapping of cutting and ground skidding

in group work

MOHAMMAD REZA GHAFFARIAN, KARL STAMPFER, JOHN SESSIONSForwarding productivity in Southern Austria

JANUSZ SOWA, DARIUSZ KULAK, GRZEGORZ SZEWCZYKCosts and efficiency of timber harvesting by NIAB 5–15 processor

mounted on a farm tractor

MOHAMMAD REZA GHAFFARIAN, HOOSHANG SOBHANIOptimization of an existing forest road network using Network 2000

Pobolj{anje vu~ne zna~ajke skidera promjenom tlaka u gumama i primjenomlanaca na kota~ima

Ocjena dvaju sustava pridobivanja drvnoga iverja iz smrekovih {uma o{te}enihpojavom potkornjaka

Utjecaj radnih uvjeta na preklapanje sje~e i privla~enja drva pri skupnom radu

Proizvodnost izvo`enja drva u ju`noj Austriji

Djelotvornost pridobivanja drva procesorom NIAB 5–15 postavljenim napoljoprivrednom traktoru

Optimizacija postoje}e mre`e {umskih cesta aplikacijom Network 2000

28 issue 2 2007 - crojfe · preventing slip, soil compaction and sinkage, and ... w – dynamic...

Documents