Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=tife20

Download by: [Luke Natural Resources Institute Finland] Date: 16 December 2015, At: 23:02

International Journal of Forest Engineering

ISSN: 1494-2119 (Print) 1913-2220 (Online) Journal homepage: http://www.tandfonline.com/loi/tife20

Productivity and work processes of small-treebundler Fixteri FX15a in energy wood harvestingfrom early pine dominated thinnings

Yrjö Nuutinen & Rolf Björheden

To cite this article: Yrjö Nuutinen & Rolf Björheden (2015): Productivity and work processesof small-tree bundler Fixteri FX15a in energy wood harvesting from early pine dominatedthinnings, International Journal of Forest Engineering, DOI: 10.1080/14942119.2015.1109175

To link to this article: http://dx.doi.org/10.1080/14942119.2015.1109175

Published online: 16 Dec 2015.

Submit your article to this journal

View related articles

View Crossmark data

Productivity and work processes of small-tree bundler Fixteri FX15a inenergy wood harvesting from early pine dominated thinnings

Yrjö Nuutinen a and Rolf Björhedenb

aNatural Resources Institute Finland, Joensuu, Finland; bSkogforsk, Uppsala Science Park, Uppsala, Sweden

(Received 27 February 2015; final version accepted 13 October 2015)

First thinnings are often neglected in Europe due to high harvesting costs. The studied small-tree bundler(Fixteri FX15a) was developed in order to rationalize the integrated harvesting of small-diameter energy woodand pulpwood in thinning operations, and to reduce transportation costs through load compaction. For thetime-and-motion study, three stand types were chosen, all dominated by Scots pine (Pinus sylvestris). Two weredense, small dimension stands. The third stand was a “normal” first thinning stand. The machine fells andaccumulates small-trees, which are fed into the bundling unit, where crosscutting, compaction, winding,scaling, and output of bundles of approximately 0.6 m3 solid is performed in an automated process. Theproductivity in the dense stand with rich undergrowth was 9.7 m3/PMh (productive machine hour), with anaverage tree volume of 27 dm3 and 3216 trees felled per hectare. In the dense stand with no undergrowth, aproductivity of 11.9 m3/PMh was reached. The average tree volume here was 44 dm3, and 2019 trees perhectare were harvested. In the normal first thinning stand with no undergrowth, the felled trees averaged 84dm3, 1266 trees per hectare were felled, and a productivity of 13.8 m3/PMh was registered. When comparedwith the previous version, the Fixteri II, the productivity of Fixteri FX15a was 2.1–2.3 times higher, dependingon the density and size of the removal. Several factors may explain the increased productivity, but one of themost prominent is an improved technical potential for multi-tree handling.

Keywords: bundling; dense stands; Fixteri; productivity

Introduction

First thinnings constitute an operational dilemma inEurope: they should be carried out to improve thedevelopment of the future stand, but are oftenneglected due to high harvesting costs. The causesof the high harvesting costs are the small stem sizeand low removals. In addition, dense undergrowthoften weakens the profitability of early thinnings(Kärhä et al. 2006; Oikari et al. 2010). The cuttingcosts constitute almost half of the production costsof forest chips from small-diameter (DBH < 10 cm)whole-trees (Kärhä et al. 2006; Laitila 2008). Inharvesting the small-trees, the biggest factor is fell-ing-piling, which accounts for 60–75% of the costs(Iwarsson Wide 2010). Based on previous studies(e.g. Brunberg et al. 1990; Sluss 1991; Johansson& Gullberg 2002; Kärhä et al. 2004, 2005;Nurminen et al. 2006; Nuutinen et al. 2010),Belbo (2011) states the strong positive influence of

stem size on the harvesting costs: “The hourly cost ofthe man-machine system is almost independent ofthe tree size, while the efficiency of the harvesting ishighly dependent.”

According to Oikari et al. (2010), the cost effi-ciency of small-diameter energy wood harvestingcould be increased by improving harvesting condi-tions (e.g. delaying harvesting operations and pre-clearance of dense undergrowth), by rationalizingharvesting methods (e.g. integrated pulpwood andenergy wood harvesting), and by training forestmachine operators.

Machine innovations and novel harvesting andhauling methods are other ways to decrease woodprocurement costs of early thinnings. In the Nordiccountries several different cutting techniques havebeen launched, during the last decade, to increaseharvesting productivity in young stands (Bergström2009; Belbo 2011; Laitila 2012). So far, the most

Corresponding author. Yrjö Nuutinen, Natural Resources Institute Finland, Production systems, P.O. Box 68, FI-80101Joensuu, Finland. Email: yrjo.nuutinen@luke.fi

International Journal of Forest Engineering, 2015http://dx.doi.org/10.1080/14942119.2015.1109175

© 2015 Forest Products Society

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

successful small diameter wood thinning supplychain has been multi-tree harvesting with a two-machine chain. In this system, a harvester or feller-buncher cuts the trees, and a forwarder hauls themto the roadside storage (Kärhä et al. 2005; IwarssonWide 2010). In multi-tree handling, the accumulat-ing felling head is capable of felling and bunchingseveral trees in each boom cycle (Bergström 2009;Iwarsson Wide 2010; Belbo 2011; Nuutinen et al.2011).

In 2008, the EU announced the 20/20/20 targets(European Commission 2008). The targets call for areduction in EU greenhouse gas emissions of at least20% below 1990 levels, with 20% of EU energyconsumption to come from renewable resourcesand a 20% reduction in primary energy use. Formany European countries, forest biomass for energyplays an important role in achieving the nationaltargets in the climate and energy strategy of the EU(Mantau et al. 2010; Bergström & Matisons 2014).For example, forest chip production for energy inFinland has increased from less than 2 TWh in theyear 2000 to 15 TWh in 2011, with 1 TWh corre-sponding to approximately 0.5 million m3 (Ylitalo2012). In Sweden, the use of primary forest fuels(obtained directly from the forest) has decreasedslightly and was 20 TWh in 2013 (Bergström &Matisons 2014; Brunberg 2014).

The studied whole-tree bundler was developed inFinland by Fixteri Oy in order to rationalize theintegrated harvesting of small-diameter energywood and pulpwood in thinning operations, and toreduce transportation costs through load compac-tion. Whole-tree bundling in early thinning can gen-erate significant benefits for the procurement system,by increasing the productivity of forwarding andmaking long-distance transport economically feasi-ble (Kärhä et al. 2009; Nuutinen et al. 2011). In thedeepest mode of integration, the pulp and energyfractions may be separated in the debarking drumof the pulp mill (Kärhä et al. 2009; Jylhä et al. 2010).

The first prototype of the studied whole-treebundler was launched in 2007 by Biotukki Oy(Jylhä & Laitila 2007). In the study by Jylhä andLaitila (2007), the productivity of this prototypewas relatively low when compared to other availableharvesting technology of whole-trees. However, theconcept showed potential for further development.In 2007–2009, a second version (Fixteri II) wasconstructed by Fixteri Oy (Kärhä et al. 2009;Nuutinen et al. 2011).

Compared to the first prototype of the whole-treebundler (Fixteri I), the productivity of Fixteri II was38–77% higher, depending on the stand density andmean tree volume of the removal. The higher

productivity level of Fixteri II was mainly a resultof increased multi-tree felling and the introductionof grapple feeding of the whole-tree bunches.Furthermore, the improved hydraulic capacity ofthe base machine enabled more simultaneous work-ing processes. Although the productivity of the sec-ond whole-tree bundler version was significantlyhigher than that of the first prototype, further devel-opment was still required. The wood supply systembased on whole-tree bundling was not yet econom-ical, as the cost savings in forwarding, long-distancetransport, and integrated comminution did not coverthe additional cost caused by compacting whole-trees into bundles (Kärhä et al. 2009; Jylhä et al.2010; Nuutinen et al. 2011). In 2012, a thirdmodel whole-tree bundler, Fixteri FX15a, waslaunched (Figure 1).

The objective of this study was to evaluate theproductivity level of this third model, Fixteri FX15a,in small-tree harvesting for different stand types. Thespecific objectives were to investigate the felling-feeding sequence and the functions of the bundlingunit and to provide suggestions for further develop-ment of the concept.

Materials and methods

The Fixteri FX15a small-tree bundler and itswork process

The studied whole-tree bundler consists of aLogman 811FC base machine, a Nisula 280E+accumulating felling head, and a Fixteri FX15abundling unit (Figure 1). The small-tree bundler is

Figure 1. The studied whole-tree bundler concept:Logman 811FC base machine, Nisula 280E+ accumulat-ing felling head, and Fixteri FX15a bundling unit (Photo:Rolf Björheden).

2 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

890 cm long, 290 cm wide and 390 cm high. Its totalweight (including the bundling unit of 6500 kg) was23,500 kg. The bundling unit is 410 cm long,240 cm wide and 280 cm high. The power for theelectric and hydraulic systems of the bundler is sup-plied from the base machine (Fixteri Ltd. 2014).

The operation consists of two main processes:the whole-tree felling/feeding (=cutting process)and the subsequent bucking and compaction oftrees into bundles (=bundling process) (Figure 2,Table 1). The Fixteri fells and feeds trees onto thefeeding table of the bundling unit, where rollers feedthe whole-trees into the feeding chamber. Then, theguillotine installed at the chamber gate bucks thetrees in the feeding chamber into 2.6 m lengths.Next, the cut trees are lifted from the feeding cham-ber into the central chamber. When there is enoughtree sections for one bundle, approximately450–550 kg green mass, the undelimbed tree

sections are elevated into the compaction chamber,where they are compressed and bound with cord ornetting. Most of the bundling operation is auto-matic, enabling simultaneous felling during bund-ling. When ejected from the bundling chamber,each bundle is weighed before it is dropped on theground. The bundles are later forwarded andstacked at a roadside landing.

Time-and-motion study of cutting (sub-study 1)

The productivity level of the new Fixteri FX15a systemin small-tree harvesting was studied in three differentstand types. The specific objectives were to investigatethe felling-feeding sequence and the functions of thebundling unit and to provide suggestions for furtherdevelopment of the concept. The influence of treevolume (average whole-tree volume of harvested trees)on the operational efficiency was analyzed. The average

Figure 2. Flowchart describing the work processes divided into work elements, for the study on the time consumption byoperation for the whole-tree bundler.

International Journal of Forest Engineering 3

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

whole-tree volume (dm3) is the total above-ground solidaverage volume of harvested trees per time study plot.

A time-and-motion study was carried out inCentral Finland in late winter in March 2013(Björheden & Nuutinen 2014; Nuutinen &

Björheden 2014). In modern usage, time-and-motion studies cover broad and practical applica-tions, combining the time study work of Taylor andthe motion study work of Gilbreth (Niebel 1988;Nuutinen 2013). Time study covers all the ways inwhich time consumption is measured and analyzedin work situations (Groover 2007). The purpose ofmotion study is to describe the work motions and toreduce ineffective movements (Niebel 1988;Groover 2007). Data were collected from nine timestudy plots located in three separate stands. Theproductive working time with all delays excluded ofeach time study plot was about 1 hour. In eachstand, three time study plots were established.To investigate the effects of stand density, under-growth and tree size on productivity the plots in thestands were selected to represent the following standtypes:

● dense, pine dominated, with rich under-growth, removal 6–9 cm DBH;

● dense, pine dominated, with little or no under-growth, removal 6–9 cm DBH;

● normal pine dominated first thinning, removal10–14 cm DBH.

Time study plots were 50 m long and 20 m wideand included two rectangular stand data plots of100 m2 from which stand data were collected(Figure 3). Table 2 presents the stand characteristicsof the time study plots.

The output was recorded through the whole-treebundler’s on-board production statistics system(time and weight of each bundle, as recorded bythe computer) in the three stand types, as the num-ber of bundles per time study plot and m3 per pro-ductive machine hour (m3/PMh) with all delaysexcluded. In the Nordic countries PMh is to equiva-lent to Effective time, E0h or G0h, (Nordic ForestWork Study Council 1978). The bundle weightswere transformed into solid cubic meters using theconversion factor 855 kg = 1 m3 (Lindblad et al.2010a, 2010b; Lindblad 2013). No samples were

Table 1. Divisions of work elements used for the time-and-motion study.

Work element Priority Definition

Crane out 1 Starts when the boom begins tomove toward a tree and endswhen the felling cut begins.

Fell A/B 1 Starts when the felling cut ofthe first tree of the grapplebunch begins and ends afterfelling and accumulating thelast tree of the grapple bunch.Fell A: trees ≥ 6 cm DBHand Fell B: trees < 6 cmDBH; the number of trees ineach grapple bunch isrecorded.

Crane in 1 Transferring the bunch of treesto the bundling unit: startsafter felling cut of the last treeof the bunch and ends whenthe bunch is on the feedingtable of the bundling unit.

Feed 1 Feeding the trees into thefeeding chamber: beginswhen the feed rollers start topull the trees on the feedingtable of the bundling unit andends when the trees are in thefeeding chamber.

Arrangement oftrees

1 Arrangement of felled trees:starts when the boom beginsto move toward a tree andends when the next workelement begins.

Arrangement ofbundles

1 Arrangement of producedbundles: starts when theboom begins to move towarda bundle and ends when thenext work element begins.

Moving 2 The bundler moves forward orbackward from one workinglocation to another: startswhen the wheels are turningand ends when the wheelsstop.

Bundling 3 All bundling operations of thebundling unit.

Dropping/weighing abundle

3 Starts when weighing thefinished bundle begins andends when the bundle isdropped on the left side ofthe strip road.

Delays(interruption)

Technical or organizationalinterruption for the whole-tree bundler’s operation.

Figure 3. Time study plot including stand data plots.

4 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

collected to define the volume conversion factor.In the study by Kärhä et al. (2009), the volumeconversion factor varied from 809–988, with anaverage of 906 and the standard error 41. In thisstudy, the dry masses of whole-trees were trans-formed into fresh weights using a moisture contentof 55%. In the study by Kärhä et al. (2009), themoisture content of fresh bundles varied in therange of 53–57%.

The time-and-motion study of cutting involvedone researcher observing the work, using a handheldfield computer (Rufco DL 2). Time was recordedusing the continuous timing method, and work ele-ments were separated from each other by numericcodes. The work processes were recorded anddivided into work elements according to Table 1.The time study data was recorded by the first author,a professional work study researcher with severalyears of experience on time studies of mechanizedlogging (Nuutinen et al. 2008; Nuutinen 2013).During the experiment, the researcher observed thework productivity outside the risk zone, so that hewas not disturbing the work of the operator.Furthermore, the observer had all the work safetyequipment according to industrial safety legislation:an orange safety vest and a safety helmet with hear-ing protectors.

The main purpose of the time-and-motion studyof cutting was to evaluate the felling-feedingsequence of the Fixteri. When recording the workelements, the crane functions had the highest priority,and the moving and bundling elements were next inpriority, respectively (Table 1). In other words, dur-ing crane functions, other simultaneous work ele-ments were not recorded.

In the experiment of this study, the operationconsisted of the felling of whole-trees and the buck-ing-compaction of whole-trees into energy wood

bundles. The time study experiments were con-ducted in a traditional strip road thinning method(Figure 3). During the experiments, the smallesttrees (DBH <6 cm) were harvested and compactedinto bundles only when they were included at noextra time expenditure, meaning together with largertrees, otherwise they were left in the stand.

It is a well-known fact that the operator,machine, and environment have a substantial influ-ence on the general work output, particularly inmechanized logging (Kariniemi 2006; Väätäinenet al. 2005; Ovaskainen 2009; Palander et al.2012). In the present study, the terrain conditionswere equal on all time study plots. Using only oneskilled operator under comparable conditionsstrengthens the reliability of the results regardingthe effects of stand density and tree size onproductivity.

Technical function test (sub-study 2)

The studied Fixteri whole-tree bundler concept con-sists of three sub-systems:

(1) A machine platform (chassis), providingpower (hydraulics) and mobility;

(2) A felling-feeding system (crane and accumu-lating felling head);

(3) A bundling system (the Fixteri bundlingunit).

As part of the evaluation a technical function testwas carried out, to investigate if bottlenecks betweenthe sub-systems 2 and 3 were restricting overall pro-ductivity (Björheden & Nuutinen 2014; Nuutinen &Björheden 2014). The technical function test wasbased on the information in Table 3. There, thetotal above-ground (= whole-tree, including

Table 2. Stand characteristics of the time study plots.

Standmeasurement Stand characteristics

Young dense richundergrowth

Young dense noundergrowth

Normal firstthinning

Stand beforecutting

Trees/hectare 4033 2836 1999Age, years 25–30 25–30 35–40

Removal,whole-trees

Removal,trees/hectare

3216 2019 1266

Average tree, dm3 27 44 84Stand aftercutting

Trees/hectare 817 817 733Solid stem volume,m3/hectarea

57 45 76

Pine/Spruce/Birch %b 73/2/24 100/0/0 86/5/9

Notes: awithout top and branches; bthe tree species mixture-%, of number of trees.

International Journal of Forest Engineering 5

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

pre-commercial top and branches) fresh weights ofremoved trees is presented, based on the functions ofRepola (2009), with an assumed moisture content of55%. Tree heights were computed using the modelof Siipilehto (1999). The fresh weights were con-verted into solid volumes using the conversion factor855 kg = 1 m3 (Lindblad et al. 2010a, 2010b;Lindblad 2013).

Capacity of the crane work

For the crane work, the productivity analysis wasconducted for the range of removed whole-treesizes that the studied felling head was able to effec-tively handle, with stump diameters ranging from5 cm (= 4.50–5.49 cm) to 17 cm (Table 3). Forthe analysis, ideally sized grapple bunches wereconstructed. An ideal bunch was defined as themaximum number of trees of a certain size that fitsin the felling head. The felling head’s maximumcapacity (Table 3) was defined by the manufacturerNisula Forest Ltd.

The maximal number of trees in the felling headis a theoretical mean value, which means that inpractice, the number of trees in the full felling headmay differ from these values. The reason for this isthe variation of branches, curves, and lengths in thetrees. In addition, the position of the tree in the headand the operator’s working technique play importantroles for the proportion of multi-tree handled trees.According to the manufacturer, the studied fellinghead is not able to effectively fell and feed treeslarger than stump diameter 17 cm.

The crane cycle time model (Equation 1,Table 4) was formulated through regression analysis.The time consumption of felling-feeding (sum ofwork elements: crane out, fell A/B, crane in, feed)(Table 1) was the dependent variable, and thenumber of accumulated trees in a full felling headwas the independent variable. The time function wasbased on the actual data of the time-and-motion

study from the study stand “young, dense, noundergrowth”.

Capacity of the bundling unit

For the bundling unit analysis, in September 2013,in the region of Kuhmoinen, a second field experi-ment was conducted wherein the studied bundleharvester’s operation was video-recorded forapproximately 1 hour. From the video, the workelements of the bundling processes (see Figures 1and 2) were determined. The durations and propor-tions of the work elements were also analyzed. Thevideo analysis gave an overall picture of the bundlingunit’s function for the further detailed simulation ofthe bundling unit operation.

For the technical function test, an ideal proces-sing rate was simulated for the bundling unit. Inactual operation, the bundling unit gets the powerfor its electrical and hydraulic system from thebase machine, and the bundling capacity is relianton the power fed by the base machine.

The simulation was carried out using the Excelsoftware-based technical function matrix of thebundling unit provided by Fixteri Ltd. The bundlingunit’s ideal processing rate was simulated for thesame sizes of whole-trees as in the crane work ana-lysis (Table 3). For each tree size, a goal weight

Table 3. The features of whole-trees used in the technical function test.

Stumpdiameter(cm)

Breast heightdiameter (cm)

Height(m)

Total above-ground solidvolume (dm3)

Total above-groundfresh weight (kg)

Number of trees in fullgrapple bunch (n)

5 2.4 3.9 3.0 2.6 117 4.0 5.9 7.5 6.4 79 5.6 7.5 15.4 13.2 511 7.2 8.8 27.5 23.5 313 8.8 9.9 44.4 38.0 215 10.4 10.7 66.3 56.7 117 12.0 11.5 93.2 79.7 1

Table 4. Statistical information on the regression modelfor the time consumption of cutting (tc), with time con-sumption of cutting (in seconds) being the dependentvariable and the natural logarithm of the number of accu-mulated trees (Ln(ntrees) in the full felling head being theindependent variable (Equation 1, Figure 9). The model isbased on 171 observations (n), and yielded an r2 of 0.252.

VariableParameterestimate SE p-value

Intercept 20.946 1.652 <0.0001Ln(ntrees) 10.982 1.456 <0.0001

6 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

bundle was defined, being as close as possible to500 kg (approximately 0.6 m3 solid), with an integernumber of ideally sized grapple bunches (Table 3).In the simulation, the ideally sized grapple buncheswere assumed to feed without interruption into thebundling unit. For each tree size class, an ideal timefor processing the goal bundle was simulated. Theideal time is the shortest technically possible timeneeded to perform the sum of work elements of thebundling process (feeding, cross-cutting, lifting, andcompacting the tree bunches to achieve a goal weightbundle). In the simulation, the variables for each treesize class were:

● goal weights of the bundles for each tree sizeclass;

● the stump diameter, length, and solid volumeof a tree in a specified tree size class;

● the number of trees in an ideally sized grapplebunch for each tree size class;

● the number of ideally sized grapple bunchesneeded to achieve a goal bundle.

In the simulation, the numbers represent an idealsituation that is not likely to occur in real operations.However, this is of minor importance for this part ofthe study, since the aim of this analysis is to investi-gate the balance between the two interdependentwork processes of felling-feeding and bucking-bund-ling, respectively.

Results

Distribution of effective work time

According to the observations of sub-study 1(Table 1), the operation of the whole-tree bundlerconsists of three main work processes:

(1) The work process felling-feeding includes cut-ting the whole-trees and bringing them to thebundling unit (work elements: crane out, fellA/B, crane in, and feed);

(2) The work process arrangement of productsincludes sorting the felled trees and pro-cessed bundles on the ground (work ele-ments: arrangement of trees andarrangement of bundles);

(3) The work process operation excluding cranework includes moving during the operation,cutting the whole-trees in the bundling unit,compressing and binding the bunch of trees,weighing, and dropping the bundle onto thestrip road (work elements: moving, bundling,and dropping/weighing a bundle).

When recording the entire working process, fell-ing-feeding and arrangement of products had the high-est priority. Work elements of these work processeswere conducted using the harvester crane. The thirdprocess, operation excluding crane functions, consistingof the work elements moving, bundling, and drop-ping/weighing a bundle, was next in priority(Table 1). Occasionally, the work elements moving,bundling, and dropping/weighing a bundle occurredsimultaneously with the crane functions (felling-feeding and arrangement of products). Such over-lapping durations were not recorded.

The proportion of the work process felling-feed-ing decreased from 83 to 61% as the average volumeof the removed trees increased from 27 to 84 dm3

(Figure 4, Table 2).

Productivity

The observed productive machine hour, PMh, was9.7 m3 in the dense young stand with rich under-growth (removal 3216 trees per hectare with anaverage whole-tree volume of 27 dm3). In thedense stand with no undergrowth, a productivity of11.9 m3/PMh was reached (removal 2019 trees perhectare, 44 dm3/tree). In the first thinning, at anaverage whole-tree volume of 84 dm3 and a removalof 1266 trees per hectare, the productivity was13.8 m3/PMh (Figure 5).

The work process felling-feeding averaged10.6 seconds per tree, with 8.2 seconds per tree inthe dense stand with rich undergrowth, 10.1 secondsper tree in the dense stand with no undergrowth, and13.4 seconds per tree in the first thinning stand.

Figure 4. Relative time consumption for the main workprocesses of the whole-tree bundler operation.

International Journal of Forest Engineering 7

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

For the previous prototype, Fixteri II, in the studiesof Kärhä et al. (2009) and Nuutinen et al. (2011),the same work process averaged 16.3 seconds pertree (Figure 6), at an average tree size of 40 dm3.

On average, 4.3 trees were accumulated percrane cycle in the dense stand with rich under-growth, 3.3 trees in the stand with no undergrowthand 2.1 trees in the first thinning stand. An averageof 2.9 trees were accumulated per crane cycle by the

previous prototype, Fixteri II (Kärhä et al. 2009;Nuutinen et al. 2011) (Figure 7).

The distribution of crane cycles by degree ofaccumulation (Figure 8) demonstrates the technicalsuitability of the equipment and the applied workingmethod in the study stands. The proportion of grap-ple bunches with two or more trees was 90% for thedense stand with undergrowth, 94% for the standwith no undergrowth, and 68% for the first thinningstand. The values indicate that the undergrowthoccasionally delimited the possibility to accumulate.The corresponding proportions of grapple buncheswith three or more trees were 78, 68, and 31%. It islogical that the tree size is more of a determiningfactor in multi-tree handling, when more than twotrees are accumulated. In the study stands, the newprototype reached an average proportion of 84% ofcrane cycles including multi-tree handling(Figure 8). The average proportion of multi-tree fell-ing with the second prototype was 80% (Kärhä et al.2009; Nuutinen et al. 2011).

Technical function test

Capacity of the crane work

The productivity of felling-feeding depends on thesize of the grapple bunch, and on the time for mov-ing and positioning the head to fell the trees includedin the grapple bunch, the time for felling the indivi-dual trees, and the time for feeding the ready grapplebunch into the bundle unit. The size of the grapplebunch is the sum of the volumes of the individualtrees in a grapple bunch. The cycle time for felling-feeding was predicted on the basis of the observed

Figure 5. Productivity of the whole-tree bundler in thethree different study stands.

Figure 6. The average time consumption for the workprocess felling-feeding in the three different study standsand for the previous prototype, Fixteri II.

Figure 7. The average number of trees per crane cycle inthe three different study stands and for the previous pro-totype, Fixteri II.

8 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

average time for felling-feeding various numbers ofaccumulated trees, expressed by the function(Equation 1, Figure 9, Tables 1, 4):

tc¼ 10; 982ln ntreesð Þ þ 20; 946 (1)

wheretc = the time consumption of cutting (work ele-

ments: crane out, fell A/B, crane in, feed), sntrees = the number of accumulated trees in the

full felling headnumber of grapple bunches (n) = 171r2 = coefficient of determination = 0.252Although the equation is based on a regression of

observed values (see Table 4), it represents an idealvalue, especially for the smallest tree size classes,

through the assumption that even very high numbersof trees may be accumulated without problems, untilthe basal area of the trees completely fills the accu-mulation area of the felling head. In reality, branchesand imperfections in stem form will delimit thedegree of accumulation before the maximum theore-tical basal area is reached.

The productivity of crane work increased signifi-cantly when the size of the accumulated treesincreased: when the stump diameter/size of the cuttree was 5 cm/3.0 dm3, the productivity of felling-feeding was 2.5 m3 per Effective working hour (m3/PMh). For the biggest trees, stump diameter 17 cmand size 93.2 dm3, the productivity increased to16.0 m3/PMh. The felling head was able to handlemultiple trees of up to a stump diameter of 13 cm.When the stump diameter of the trees increasedfrom 13 to 15 cm, the trees were processed indivi-dually, and in that case the productivity increasedonly from 11.2 to 11.4 m3/PMh (Figure 10).

Capacity of bundling unit

The production capacity of the bundling unit variedfrom 23.3–36.3 m3/PMh. To produce one goalweight bundle (approximately 500 kg) 198–6whole-trees were needed for stump diameter classesfrom 5–17 cm (Table 3). When the length and sizeof grapple bunch trees increased, the productivityalso increased slightly. The reason for this was thatthe duration of feeding and cutting the tree bunchesin the feeding chamber decreased when the lengthand size of the grapple bunches increased. The high-est productivity figures (36.3 m3/PMh and 29.1 m3/PMh) were reached for processing the biggest goalbundles (526.4 kg and 531.5 kg). The results indi-cate that the bundle size has a positive influence onproductivity (Figure 11).

Figure 8. Distribution of crane cycles with different numbers of accumulated trees in the study stands.

Figure 9. Time consumption of felling-feeding (workelements crane out, fell A/B, crane in, feed) as a functionof the number of trees accumulated in the felling head.

International Journal of Forest Engineering 9

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

Balance between the bundling unit and felling-feeding

The bundling unit has a significantly higher capacitythan the crane and felling head. The balancebetween these two sub-systems was analyzed by set-ting the capacity of the bundling unit to 1 and there-after comparing the respective relative capacity offelling-feeding. The analysis shows that the felling-feeding sub-system restricts the production of theconcept. The capacity of felling-feeding varied from11–57% of the capacity of the bundling dependingon tree size (Figure 12).

Discussion

The productivity of the Fixteri FX15a whole-treebundler is significantly higher than reported for theprevious prototype, Fixteri II. In the dense standwith undergrowth, the productivity for the Fixteri

FX15a was 9.7 m3/PMh, and in the dense standwith no undergrowth, a productivity of 11.9 m3/PMh was reached (Figure 5). In the studies ofFixteri II by Kärhä et al. (2009) and Nuutinenet al. (2011), the average productivity was 5.1 m3/PMh, when the density of removal was 1400 treesper hectare and the average tree volume was 40 dm3,thus under similar conditions. With an averagewhole-tree volume of 31 dm3 and a removal densityof 2850 trees per hectare, a productivity of 4.6 m3/PMh was reached. Thus, the productivity of theFixteri FX15a is 2.1–2.3 times higher, dependingon stand density and average whole-tree volume ofthe removal (Figure 13).

The higher productivity of the new Fixteri FX15aresults from increased multi-tree felling-feeding,resulting in a decrease of 35% in crane cycle timeper tree, compared to the previous Fixteri II model(Figure 6). The increased performance of crane

Figure 10. Ideal productivity of the felling-feeding system for the maximum accumulated grapple bunches in the selectedtree sizes, according to Table 3.

Figure 11. The ideal processing rate of the bundling unit for the same sizes of whole-trees as in the crane work analysis(Table 3).

10 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

work enables more effective feeding of the bundlingunit. The new felling head could also transfer moretrees per grapple bunch to the bundling unit. Theaverage number of trees per crane cycle was 4.3 treesfor the dense stand with undergrowth, 3.3 trees forthe stand with no undergrowth, and 2.1 trees for thefirst thinning stand. Respectively, on average, 2.9trees were accumulated per crane cycle by the pre-vious prototype, Fixteri II (Kärhä et al. 2009;Nuutinen et al. 2011) (Figure 7). Iwarsson Wide

(2010) states that the number of trees per cranecycle is a critical parameter for multi-tree handlingin small diameter stands. Belbo (2011) showed,through simulation, that the optimal number ofaccumulated trees in multi-tree cutting is 4–5 treesper crane cycle.

The two young dense study stands, where the aver-age volume of removedwhole-trees was 27 and 44 dm3

respectively, proved to be a more suitable operatingarea for the accumulating felling head than the firstthinning stand. There, the removed trees, averaging84 dm3, were too large for multi-tree handling. Theconclusion is based on the following observations:

● duration for the work element arrangement offelled trees (Table 1) increased from a level of3% in young stands to more than 7% in thefirst thinning; and

● the average proportion of multi-tree handledtrees (meaning grapple bunches with two ormore trees) was only 68% in the first thinning,compared to 90% and 94% in young stands.

In the technical function test, the ideal productioncapacity of the bundling unit was from 2–10 timeshigher than the possible level of felling-feeding in theworking environment of sub-study 1 (Figure 12). Theresults indicate that the studied whole-tree bundlingconcept is optimal when processing trees of an aver-age DBH of 6–9 cm. For bigger trees, felling-feedingbecame more difficult, while for smaller trees, theharvesting cost is probably too high.

The capacity of felling-feeding was mostly depen-dent on the size of the trees and on the grapplebunch size. The productivity of felling-feedingdecreased significantly for the smallest trees (stumpdiameter <9 cm, Figure 10). The studied fellinghead could handle multiple trees of up to a stumpdiameter of 13 cm. Trees from stump diameter15 cm were processed individually, indicating thataccumulation of two 15 cm trees was not practicable(Figure 10, Table 3). These findings should be con-firmed by further experiments.

In the actual time-and-motion study, for thestudy stand young dense no undergrowth, the aver-age duration of accumulation of the trees (work ele-ment: fell A/B) increased from 3.4 to 21.5 secondsper grapple bunch, when the number of accumu-lated trees increased from 1 to 5 trees.Respectively, the combined duration of the workelements crane out, crane in, and feed stayed at aconstant level, in the range of 17.6–20.4 seconds pergrapple bunch cycle (see Figure 8, Table 1). Theseresults indicate that felling-accumulation is animportant development target for multi-tree cutting.

Figure 12. The production capacity of felling-feedingwas in the range of 0.1–0.6 of the respective ideal capacityof the bundling unit, based on the technical simulationstudy.

Figure 13. Recorded productivity per average removal ofthe present study in the three study stands, compared tothe results of Fixteri II studies (Kärhä et al. 2009;Nuutinen et al. 2011).

International Journal of Forest Engineering 11

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

The results of the technical function test implythat felling-feeding is a bottleneck for the studiedmachine concept. In order to increase the perfor-mance, the efficiency of felling-feeding should beincreased, for example by:

● improved crane work by, e.g. optimal crane-geometry, cabin design, and semi-automation;

● improved function of the felling head for theaccumulation, sorting and handling of felledtrees;

● developing feeding of the tree bunches into thebundling unit;

● applying a technology that enables continuousaccumulation and felling of trees;

● especially in harvesting trees of a whole-treevolume of less than 15–20 dm3, consideringsome other technical felling-head solution asthe studied head.

The simulation of the ideal productivity of thebundling unit indicates that the best bundling pro-ductivity is reached by producing the biggest bundlespossible. This demands cutting and feeding thelongest and biggest whole-tree bunches possibleinto the bundling unit (Figure 11). The size andlength of the whole-tree bunches seem to have themost significant influence on the feeding and cuttingspeed of stems in the bundling unit. The reason forthis is that, when compacting the shorter trees, alarger number of tops that are under 2.6 m longdecreases the feeding speed of tree bunches.

Ala-Varvi and Ovaskainen (2013) studied thecompetitiveness of the Fixteri FX15a small-treebundler, compared to the cutting of undelimbedtrees with a John Deere 1070D harvester equippedwith a H412 accumulating harvester head. In thestudy, the productivity of Fixteri FX15a was9.7 m3/PMh with an average tree volume of 37dm3. Respectively a productivity of 10.5 m3/PMhwas reached in cutting of undelimbed trees with anaverage tree volume of 48 dm3.

Ala-Varvi and Ovaskainen (2013) determined thecosts of cutting and forwarding of both supply chainsat an average tree volume of 40 dm3, forwardingdistance of 300 m and harvesting removal of 60 m3

per hectare. The cost of whole-tree bundling opera-tion was 13.6 €/m3 and for cutting of undelimbedtrees 15.3 €/m3. The cost for forwarding the bundleswas 3.57 €/m3 and respectively for delimbed trees5.56 €/m3. They found the total cost of the woodchip supply chain of whole-tree bundling (46.7€/m3) to be lower compared to undelimbed trees(50.2 €/m3) if the average volume of removed treeswas less than 85 dm3.

Also Bergström et al. (2015) studied the effectof tree size and undergrowth on the operationalefficiency of the Fixteri FX15a bundle harvesterin early fuel wood thinnings in the North ofSweden. Their studies of the machine productivityfocused on stands with an average harvested treesize below 30 dm3 (the average tree volume in thetime study plots varied from 15–43 dm3). Theharvested forest was 30–35-years-old and com-posed mostly of Scots pine, Norway spruce andbirch. The recorded performance is similar to thefindings of this study.

In the dissertations of Bergström (2009), Belbo(2011), Jylhä (2011), Röser (2012), Laitila (2012),and Nuutinen (2013), the feasibility of methods andtechniques in pre-commercial and early thinning wasevaluated and developed. The studies aimed toincrease the understanding of the functions of har-vesting work, develop the efficiency of supply chainsand estimate their costs, evaluate existing technologyand suggest improvements for reducing costs andevaluate the effects of using new methods and tech-niques intended to increase the efficiency. Thesestudies have confirmed that new techniques andmethods can significantly increase the productivityand cost-efficiency of small-diameter harvesting.

Several studies have shown that the operator hasa significant effect on harvesting productivity (e.g.Kariniemi 2006; Sirén 1998; Ryynänen & Rönkkö2001; Laamanen 2004; Väätäinen et al. 2005;Purfürst & Erler 2006). In the study by Väätäinenet al. (2005), the differences between the harvesteroperators in thinning productivity varied from 40–55%, depending on the stem size processed. Theproductivity leap of the new Fixteri model recordedin the present study, from 111 to 133% compared tothe previous model, is so significant that, althoughthe operator was not the same as in previous studies,the new technical features of the whole-tree bundlerare likely to account for a substantial part of theincreased productivity.

The findings of this and other recent studies (Ala-Varvi & Ovaskainen 2013; Bergström et al. 2015) ofthe Fixteri FX15a indicate that bundling whole-treesis an interesting alternative for harvesting small-dia-meter trees. The competitiveness is highly dependenton wood chip prices, energy and forest policies etc. aswell as on the characteristics of the stand. Furtherstudies of the bundling operation in varying harvest-ing conditions and methods are desirable.

Disclosure StatementNo potential conflict of interest was reported by theauthors.

12 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

ORCIDYrjö Nuutinen http://orcid.org/0000-0003-3360-4444

ReferencesAla-Varvi T, Ovaskainen H. 2013. Kokopuun paalauksen

kilpailukyky [Internet] [Competitiveness of whole-treebundling]. Metsäteho report 225. ISSN 1796-2374; 77p; [cited 2015 Jul 23]. Available from: http://www.metsateho.fi/wp-content/uploads/2015/02/Raportti_225_Kokopuun_paalauksen_kilpailukyky_ta-v_ho.pdf.[In Finnish with English summary].

Belbo H. 2011. Efficiency of accumulating felling heads andharvesting heads in mechanized thinning of small dia-meter trees [dissertation]. Linnaeus UniversityDissertations No. 66/2011. School of Engineering; 42 p.

Bergström D. 2009. Techniques and systems for boomcorridor thinning in young dense forests[Dissertation]. Swedish University of AgriculturalSciences, Faculty of Forest Sciences. No. 2009:87;53 p.

Bergström D, Fulvio F, Nuutinen Y, Jonnson L. 2015.Demo report 15 – Studies and demonstration on theuse of a bundle-harvester system in early fuel woodthinnings - D4.5 [Internet]. In: INFRES - Innovativeand effective technology and logistics for forest resi-dual biomass supply in the EU (311881). IVALSA,Sesto Fiorentino; 14 p; [cited 2015 Jul 23]. Availablefrom: http://www.infres.eu/en/results/?PHPSESSID=a3366a8deaf708a5c64ef8f16155fe23

Bergström D, Matisons M, editors. 2014. Forest Refine,2012–2014 –Efficient forest biomass supply chainmanagement for biorefineries, Synthesis report.Swedish University of Agricultural Sciences (SLU).Report 18; 113 p.

Björheden R, Nuutinen Y. 2014. Studie av Fixteri FX15aklenträdsbuntare [A study of small-tree bundler FixteriFX15a]. Arbetsrapport från Skogforsk nr. 819-2014;20 p. [In Swedish with English summary].

Brunberg T. 2014. Skogsbränslets metoder, sortiment ochkostnader 2013 [Internet]. [Methods, assortments andcosts of forest chip production 2013]; [cited 2014 Dec11]. Available from: www.skogforsk.se. [In Swedish].

Brunberg B, Hedenberg Ö, Jonsson T. 1990.Flerträdsteknik – effekter på avverkningskostnaderoch massaindustrins råvara [Multi-tree technology –

Its Impact on Logging Costs and Pulpmill RawMaterials]. Redogörelse nr. 4. Kista, Sweden:Forskningsstiftelsen Skogsarbeten. [In Swedish withEnglish summary].

European Commission. 2008. Combating climate change.The EU leads the way. Europe on the move series2008. European Commission. Brussels, Belgium; 24 p.

Fixteri Ltd. [Internet]. 2014. [cited 2013 Jun]. Availablefrom: http://www.fixteri.fi/

Groover MP. 2007. Work systems and the methods, mea-surement, and management of work; 778 p.

Iwarsson Wide M. 2010. Technology and methods forlogging in young stands. In: Thorsén Å, BjörhedenR, Eliasson L, editors. Efficient forest fuels supplysystems. Uppsala Science Park, Sweden: Skogforsk.

Johansson J, Gullberg T. 2002. Multiple tree handling inthe selective felling and bunching of small-trees indense stands. Int J For Eng. 13:365–372.

Jylhä P. 2011. Harvesting undelimbed Scots Pine (Pinussylvesteris L.) from first thinning for integrated produc-tion of kraft pulp and energy [dissertation]. Universityof Helsinki. Department of Forest Sciences.Dissertationes Forestales 133; 73 p.

Jylhä P, Dahl O, Laitila J, Kärhä K. 2010. The effect ofsupply system on the wood paying capability of a kraftpulp mill using Scots pine harvested from first thin-nings. Silva Fennica. 44:695–714.

Jylhä P, Laitila J. 2007. Energy wood and pulpwood har-vesting from young stands using a prototype whole-tree bundler. Silva Fennica. 41:763–779.

Kärhä K, Jouhiaho A, Mutikainen A, Mattila S. 2005.Mechanized energy wood harvesting from early thin-nings. Int J For Eng. 16:15–26.

Kärhä K, Keskinen S, Liikkanen R, Lindroos J. 2006.Kokopuun korjuu nuorista metsistä [Whole-tree har-vesting in early thinnings] [Internet]. MetsätehoReport 193; 79 p; [updated 2006 Sep 12; cited 2010Jun 23]. Available from: http://www.metsateho.fi/kategoria/raportti/page/9/. [In Finnish].

Kärhä K, Laitila J, Jylhä P, Nuutinen Y, Keskinen S. 2009.Kokopuun paalaus -tuotantoketjun tuottavuus ja kus-tannukset [Productivity and costs of the whole-treebundling supply chain]. Metsäteho Report 211. ISSN1459-773X; 60 p. [cited 2015 Jul 23]. Available from:http://www.metsateho.fi/wp-content/uploads/2015/02/Raportti_211_Kokopuun_paalaus_tuotantoketjun_tuottavuus_ja_kustannukset_kk.pdf [In Finnish].

Kärhä K, Rönkkö E, Gumse S-I. 2004. Productivity andcutting costs of thinning harvesters. Int J For Eng.15:43–56.

Kariniemi A. 2006. Kuljettajakeskeinen hakkuukonetyönmalli – työn suorituksen kognitiivinen tarkastelu[Operator-specific model for mechanical harvesting –

cognitive approach to work productivity] [disserta-tion]. Helsingin yliopiston metsävarojen käytön laitok-sen julkaisuja 38.; 126 p. [In Finnish with summary inEnglish].

Laamanen V. 2004. Hakkuukoneenkuljettajan hiljaisen tie-don näkyväksi tekeminen simulaattorin ja matemaattis-ten menetelmien avulla [Describing harvesteroperator’s tacit knowledge by simulator and mathema-tical models]. Diplomityö. Tampereen teknillinen ylio-pisto. Automaatiotyötekniikan koulutusohjelma; 69 p.[In Finnish].

Laitila J. 2008. Harvesting technology and the cost of fuelchips from early thinnings in Finland. Silva Fennica.42:267–283.

Laitila J. 2012. Methodology for choice of harvesting sys-tem for energy wood from early thinning [dissertation].University of Eastern Finland. Faculty of Science andForestry. Dissertationes Forestales 143; 68 p.

Lindblad J. 2013. Energiapuun mittauksen kehittäminen –

hankkeen loppuraportti [Developing measurement ofenergy wood –final report]; 12 p. [In Finnish].

Lindblad J, Äijälä O, Koistinen A. 2010a. Energiapuunmittaus [Energy wood measurement]; 31 p. [InFinnish].

Lindblad J, Äijälä O, Koistinen A. 2010b. Mätning avenergived. [Energy wood measurement]; 31 p. [InSwedish].

Mantau H, Saal U, Prins K, Steierer F, Lindner M,Verkerk H, Eggers J, Leek N, Oldenburger J,Asikainen A, Anttila P. 2010. EUwood – real potential

International Journal of Forest Engineering 13

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015

for changes in growth and use of EU forests. Finalreport. University of Hamburg. Centre of WoodScience. Hamburg, Germany, 160 p.

Niebel BW. 1988. Motion and time study. 8th ed.Homewood, IL: Irwin. 799 p.

Nordic Forest Work Study Council. 1978. Forest WorkStudy Nomenclature.

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

Nuutinen Y. 2013. Possibilities to use automatic and man-ual timing in time studies on harvester operations [dis-sertation]. University of Eastern Finland. Faculty ofScience and Forestry. Dissertationes Forestales 156;68 p.

Nuutinen Y, Björheden R. 2014. Fixteri FX15a –koko-puupaalaimen tuottavuus ja työprosessit nuortenmäntyvaltaisten metsien energiapuun korjuussa[Productivity and work processes of small-tree bundlerFixteri FX15a in energy wood harvesting from earlypine dominated thinnings]. Metlan työraportteja 281/Working Papers of the Finnish Forest ResearchInstitute 281; 21 p. [In Finnish].

Nuutinen Y, Kärhä K, Laitila J, Jylhä P, Keskinen S. 2011.Productivity of whole-tree bundler in energy wood andpulpwood harvesting from early thinnings. Scand J ForRes. 26:329–338.

Nuutinen Y, Väätäinen K, Asikainen A, Prinz R,Heinonen J. 2010. Operational efficiency and damageto sawlogs by feed rollers of the harvester head. SilvaFennica. 44:121–139.

Nuutinen Y, Väätäinen K, Heinonen J, Asikainen A, RöserD. 2008. The accuracy of manually recorded timestudy data for harvester operation shown via simulatorscreen. Silva Fennica. 42:63–72.

Oikari M, Kärhä K, Palander T, Pajuoja H, Ovaskainen H.2010. Analyzing the views of wood harvesting profes-sionals related to the approaches for increasing thecost-efficiency of wood harvesting from young stands.Silva Fennica. 44:481–495.

Ovaskainen H. 2009. Timber harvester operators’ workingtechnique in first thinning and the importance of cog-nitive abilities on work productivity [dissertation].University of Eastern Finland. DissertationesForestales 79; 62 p.

Palander T, Ovaskainen H, Tikkanen L. 2012. An adap-tive work study method for identifying the humanfactors that influence the Performance of a Human-machine system. For Sci. 58:377–389.

Purfürst T, Erler J. 2006. The precision of productivitymodels for the harvester – do we forget the humanfactor? In: Precision forestry plantations, Semi-naturaland natural forests. Proceedings of the internationalprecision forestry symposium; 2006 Mar 5–10; SouthAfrica: Stellenbosch University; p. 465–475.

Repola J. 2009. Biomass equations for Scots pine andNorway spruce in Finland. Silva Fennica. 43:625–647.

Röser D. 2012. Operational efficiency of forest energy supplychains in different operational environments [disserta-tion]. University of Eastern Finland, Faculty of Scienceand Forestry. Dissertationes Forestales 146; 83 p.

Ryynänen S, Rönkkö E. 2001. Harvennusharvestereidentuottavuus ja kustannukset [Productivity and expensesassociated with thinning harvesters]. Helsinki:Työtehoseuran julkaisuja 381; 67 p. [In Finnish withEnglish summary].

Siipilehto J. 1999. Improving the accuracy of predictedbasal-area diameter distribution in advanced stands bydetermining stem number. Silva Fennica. 33:281–301.

Sirén M. 1998. Hakkuukonetyö, sen korjuujälki ja puusto-vaurioiden ennustaminen [One-grip harvester opera-tion, its silvicultural result and possibilities to predicttree damage] [dissertation]. Finnish Forest ResearchInstitute. Research papers 697; 179 p. [In Finnish withEnglish summary].

Sluss R. 1991. Flerträskovning – fremtiden for de tidligetyndinger [Multi-tree handling – future for early thin-nings]. Copenhagen, Denmark: Danish Inst. of ForestTechnology. [In Danish].

Väätäinen K, Ovaskainen H, Ranta P, Ala-Fossi A. 2005.Hakkuukoneenkuljettajan hiljaisen tiedon merkityshakkuutulokseen työpistetasolla [The significance ofthe harvester operator’s tacit knowledge on cuttingwith a single grip harvester]. Joensuu, Finland:Metsäntutkimuslaitoksen tiedonantoja 937; 100 p. [InFinnish].

Ylitalo E. 2012. Puun energiakäyttö 2011 [Wood con-sumption in energy generation 2011]. Finnish ForestResearch Institute, Statistical Bulletin 16/2012. [InFinnish].

14 Y. Nuutinen and R. Björheden

Dow

nloa

ded

by [

Luk

e N

atur

al R

esou

rces

Ins

titut

e Fi

nlan

d] a

t 23:

02 1

6 D

ecem

ber

2015