Available online at www.sciencedirect.com

Biomass and Bioenergy 26 (2004) 27–37

All-year-round harvesting of short rotation coppice eucalyptuscompared with the delivered costs of biomass from more

conventional short season, harvesting systems

Ralph E. H. Simsa ;∗, Piero Venturib

aCentre for Energy Research, College of Sciences, Massey University, Private Bag 11 222, Palmerston North, New ZealandbDepartment of Economics and Agricultural Engineering, University of Bologna, Bologna 40127, Italy

Received 24 September 2002; received in revised form 4 April 2003; accepted 25 April 2003

Abstract

This study attempted to de-ne the optimum harvesting, processing and transport system in terms of the cost per tonne ofdelivering biomass produced from a commercial short rotation coppice crop to a 10 MWe bioenergy conversion plant 25 kmaway. Harvesting the crop during one short seasonal period of the year results in the need to store most of the material forbetween one to 12 months in order to provide a continual supply of feedstock. Storage of large volumes of biomass is costlyand also results in dry matter losses over time. An alternative system would be to harvest small areas as required everyfew weeks throughout the year. This would enable cheaper, lower performance equipment to be used and hence providea continual supply of biomass feedstock for delivery to the plant. Four systems of conventionally harvesting the biomassduring one short seasonal period of 8–10 weeks were compared with two systems of harvesting it continually throughout theyear. Whether the biomass is stored on the farms after harvest to allow for transpirational drying or at the plant was alsocompared. All-year-round harvesting using a simple tractor-mounted circular saw showed cost bene-ts in terms of $/GJ ofenergy delivered to the bioenergy plant but the performance assumptions made will require con-rmation by -eld monitoring.A sensitivity analysis was conducted on several key parameters.? 2003 Elsevier Ltd. All rights reserved.

Keywords: Short rotation; Coppice; Supply chain; Harvest systems; All-year-round harvest

1. Introduction

The production of short rotation forests to providefeedstock for producing heat, electricity or transportbiofuels has reached the commercial stage in sev-eral bioenergy projects [1]. However the costs of thebiomass material in terms of $/GJ delivered to the con-version plant, remain relatively high compared with

∗ Corresponding author. Tel.: +64-6-3505288; fax: +64-6-3505-604.

E-mail address: r.e.sims@massey.ac.nz (R.E.H. Sims).

coal or natural gas. These costs must therefore befurther reduced if bioenergy from these energy cropsources is to compete with fossil fuels or even withbiomass from other sources such as wood processresidues.Typical harvestable yields from short rotation

forests of 10–24 t of dry matter per hectare and yearhave been reported under warm temperate conditions[2]. Part of the delivered supply costs involve harvest-ing, transport and storage. These need to be minimisedby optimising truck payloads [3] and minimising drymatter losses when in store [4,5].

0961-9534/04/$ - see front matter ? 2003 Elsevier Ltd. All rights reserved.doi:10.1016/S0961-9534(03)00081-3

28 R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37

Typically, a short rotation forest crop is harvestedevery 3–4 years. Hence, by planting a third (or a quar-ter) of the total area in each of years 1–3 (or 4), thena regular supply of biomass can be provided as feed-stock for the conversion plant each year. Fuel supplycontracts with the growers would need to be drawn upby the developer of the conversion plant to ensure thatthe security of supply is maintained over the expectedeconomic life of the plant.The optimum time for harvesting a crop of short

rotation forest is normally a compromise betweenmaximising winter sprout production, avoiding coldwinter temperatures on young regrowth stems, avoid-ing summer dry periods and gaining maximum sum-mer growth [6]. Under conditions experienced in thecentral North Island of New Zealand, short rotationcoppice eucalyptus is normally harvested in springtime during October/November when rainfall is com-mon and the soil is warm enough to encourage plantregrowth. However, long-term -eld trials (includingsome harvested every 3 years for 15 years, yet tobe published) in the central North Island of NewZealand have shown that a stand can be harvestedall year round without the occurrence of any signif-icant agronomic problems such as increased stumpmortality or yield losses [7]. Harvesting in Januaryand February caused some yield reduction but con-versely where harvesting was delayed for severalmonths past the usual October/November harvestperiod, the additional growth which resulted wasan added advantage. This was veri-ed by Blake [8]who reported that E. obliqua showed a decline in thenumber and length of sprouts when harvested in thesummer.The objective of the study reported here was to de-

termine whether all-year-round harvesting of short ro-tation coppice eucalyptus could be a viable system ofproviding more secure and cheaper biomass fuel sup-plies compared with a single harvest operation con-ducted over a concentrated period of just 3 or 4 weeksand resulting in the biomass having to be stored forup to 12 months in large piles.All-year-round harvesting enables favourable

changes to be made to the supply chain in that:

• any change in heat and power demand with seasonscan be met by harvesting a greater or lesser areaeach week;

• a more even labour and machinery requirementwould occur throughout the year;

• additional growth would be obtained by leaving thetrees standing as long as possible before requiringthem to be harvested for the biomass;

• the storage period of the material would be greatlyreduced leading to lower dry matter losses;

• the cost of storage, whether stored under cover oron a given area of land, would be reduced as smallervolumes can be stored at any given time; and

• smaller, cheaper machines could be used with lowerperformance capacities but optimised to match theharvest requirements.

2. Methodology

In this study, the production yield data used weretaken from a trial block of E saligna planted at MasseyUniversity in November 1988, with 12 plots of singlestems harvested monthly from February 1992 untilJanuary 1993. The coppice regrowth plots were thenharvested exactly 3, 6 and 9 years later. It was assumedthat in a commercial operation, one third of the landarea would be planted in year 1 ready for -rst harvestin year 4; another third planted in year 2 and -rstharvested in year 5; and the -nal third planted in year 3and -rst harvested in year 6. Harvesting of the coppiceregrowth would then begin in year 7.It was further assumed that a 10 MWe integrated

gasi-cation, combined cycle power plant, similar tothat of ARBRE in the United Kingdom [9] was tobe constructed in the Manawatu region of the NorthIsland of New Zealand and that several short rotationforest plantations would be established around the areato supply it. Assuming the plant operated at 33% ef--ciency for 7000 h per year to meet a constant baseload, the total energy requirement would be approxi-mately 850 TJ of biomass per year.The amount of land needed to meet the constant

energy demands of the power plant was calculatedbased on a range of moisture content assumptions andyield data from the all-year-round harvest -eld trials.To meet this total energy demand, it was calculatedthat if yields of 15 oven dry tonnes per hectare andyear could be achieved on average, then a total ofapproximately 2800 ha of short rotation eucalyptusforest would need to be grown in the region. The treeswould be planted at 1:25 m spacings in 2-m-wide rows

R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37 29

Table 1Land area needed to supply 10 MWe plant based on actual harvested yields measured from -eld trials and allowing for commercial scalecropping

Harvest All-year-round harvest Seasonal harvest(ha/week) (ha/year)

Single stem -rst harvest, 3 year old 18.8 726.6First coppice harvest, 3 year regrowth 29.9 815.8Second coppice harvest, 3 year regrowth 29.1 774.3Third coppice harvest, 3 year regrowth 32.6 843.4

giving around 4000 trees per hectare. Moisture contenteKects on the lower heat value of the biomass wereassumed to be [4]

Heat value (MJ=kg)

=20− 0:214×moisture content wet basis:Yield variations between -rst, second, third and fourthharvests as measured from the -eld trials [7,10] in-cluding some unpublished data, were also taken intoaccount. The area of land needed to be harvested tomeet the annual power plant demand ranged from 18.8to 32:6 ha each week for the all-year-round harvestsystem and up to 843 ha a year for the seasonal har-vest in October/November (Table 1).Mainly to reduce the number of possible variables

of the analysis, it was assumed that in all cases thebiomass would be delivered as whole trees to thepower plant where it would be comminuted in a tubgrinder just prior to feeding the material into the -naldryer before entering the gasi-er. It was further as-sumed that the plantations, when established, wouldhave roading every kilometre suitable for heavy trucksand that every 200 m there would be a spur road andlanding area to which the harvested biomass couldbe transported from the forest and stored temporarily.The average transport distances of the biomass wouldthen be

• 200 m in the forest with a tractor and fork, for-warder or feller–buncher;

• 800 m along spur roads after loading on to atruck/trailer at an average speed of 8 km=h;

• 24 km on public A or B classi-ed highways at anaverage speed of 70 km=h.

This data was fed into a computer model whichwas used to calculate the detailed transport and stor-age costs for the various harvesting/handling systems

as outlined below. This “Biomass Transport Modelfor New Zealand” (BIOTRANZ) [3] is a simulationmodel based on that developed in the UK by EnergyTechnology Support Unit of the Department of Tradeand Industry (ETSU) [11]. It was developed to studythe harvest logistics of biomass and to identify the op-timal sequence of events in the supply chain in orderto provide fuel delivered to the power plant in the mostdesirable form and at the cheapest cost. The modelwas based on an Excel spreadsheet and used to pro-vide a comprehensive analysis and datasheet for allsystems comparing $ per tonne of dry matter and $/GJdelivered. Dry matter losses during harvesting, stor-age and processing, and changes of the moisture con-tent over time are incorporated. Previous reported useof the model has been for Salix grown in the UK [12]and forest residues from Pinus radiata plantations inNew Zealand [3].Based on data taken from New Zealand time-of-

harvest -eld trials [7] and also more recent unpub-lished data, the biomass moisture content at harvestwas taken to be 60% (wet basis) regardless of the timeof year. For the all-year-round harvested system, thematerial will be stored as whole trees at the landingfor 1 month before transporting to the power plant. Atthis stage, the moisture content would have droppedrapidly in the -rst 4 weeks to around 45% (wet ba-sis) regardless of the time of year [13]. For the sea-sonal harvest system, the whole tree material couldbe stored either on the farms or in a large pile at thepower plant. Either option would result in an averagemoisture content reduction over the 6 months aver-age storage period of around 3–4% per month as afterthe initial rapid drop, the rate of drying rapidly slowsdown. A monthly 1.5% dry matter loss was assumed,based on previous experiences under New Zealandclimatic conditions [5,6].

30 R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37

The BIOTRANZ model automatically accounts formoisture content changes over time and also for drymatter losses. To provide a -xed amount of useful en-ergy at the plant, the purchased volume of the mate-rial is increased to compensate for any losses whichresults in a higher total initial purchase price. Thusif 20 t of biomass were required and a 10% loss wasexpected prior to use in the gasi-er, then 22 t wouldneed to be purchased. In this study, the purchase priceof the biomass was set at $NZ 1 10/t dry matter.

2.1. System options compared

Six options of the two harvest chain systems underevaluation were selected for analysis as outlined inFig. 1.Options A and B were based on a simple,

tractor-mounted, circular saw harvester as devel-oped by the Italian Research Institute for Agricul-tural Mechanisation (ISMA) in Rome [14,15]. The750-mm-diameter blade was powered by the tractorhydraulics, mid-mounted on the tractor and positionedwith a variable cutting angle depending on the cropcharacteristics. It was assumed the saw would be ableto handle either single stem or multiple stem coppiceregrowth with base stem diameters up to 200 mm.The saw blade and drive mechanism is mid-mountedon the tractor when cutting, but designed to be carriedbehind the tractor during transport on the road. Theactual -eld performance of the saw (in terms of hoursper hectare) will vary with crop type and stemwoodsize and is not known. Therefore, two options wereanalysed covering an assumed work productivityrange of either 1:2 h=ha (A) or 4:5 h=ha (B).The cut stems fall at right angles to the rows and are

then gathered and collected by a tractor-mounted fork.The whole stems are then transported an average of200 m to the landing at the edge of the -eld at a workrate of 26:6 h=ha. In order to match the output of theharvester and produce a balanced system to transportthe large volumes of material, a simple spreadsheetwas developed to calculate the number of machinesneeded. The results, as shown in Table 2, were usedin the cost calculations for each system.

1 $1 New Zealand = 0:5 Euro or $0.5US approximately.

For both Options A and B it was assumed the cutstems will be left stored on the farms to dry naturallyfor one month (Fig. 2) prior to transporting them bytruck/trailer to the power plant (Fig. 1). The materialwill be unloaded by a knuckle boom grab which alsoserves to feed the -xed tub grinder selected to com-minute the material at a rate of around 34 t=h in orderto meet the conversion plant’s fuel demand.Options C, D, E and F were based on the more con-

ventional seasonal harvesting method whereby all ofthe biomass material is cut in the late winter or earlyspring during a 4 week period, then stored either onthe farm or at the conversion plant. The plant needs tobe continually supplied with fuel. So it was assumedthat some materials will be delivered directly at har-vest, some stored for only a week or two, and somestored for up to 11 or 12 months. Overall, the averagestorage period would be 6 months. The 35% (wet ba-sis) moisture content obtained after 6 months storageas whole tree piles represents an average level whichwas used throughout the study. In practice, the mois-ture content of the delivered material will vary withlength of storage and be limited by the atmospherichumidity at the time as the biomass is hygroscopic.Options C and D (Fig. 1) use a forwarder (Fig. 2)

to cut and transport the whole stems to the landingwhere they are either stored before transporting (C)or are transported directly (D) to the storage area atthe conversion plant. Similarly, Options E and F usea feller buncher and either have storage at the landingfor later collection or at the power plant.When stored at the power plant a tractor loader is

used to unload the truck/trailer and then to transfer thematerial from the large store area to the knuckle boomfeeder as required. It was assumed that this tractorloader would be unsuitable for loading the tub grinderand hence could not replace the knuckle boom feeder,partly for practical reasons and partly to keep the han-dling system as used at the power plant constant for allsix options in order to simplify the analysis by avoid-ing any further variables.

2.2. Machinery use assumptions

When the method of harvesting is chosen in prac-tice, selection of suitable equipment is based on thestand characteristics, the desired end product, thenature of the terrain, and the scale of operation [16].

R.E.H.Sims,P.Venturi/B

iomass

andBioenergy

26(2004)

27–37

31

Fig. 1. Harvest and system options compared in the study: two options for all-year-round harvesting and four options for seasonal harvest.

32 R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37

Table 2Number of machines needed to harvest the required area (allowing for wet periods and assuming some 16 h days can be worked whennecessary) and to transport and process the material to continually supply a 10 MWe power station

All-year-round Saw—high capacity Saw—low capacity Tractor fork—-eld transport Tractor fork—loaderharvest option A option B options A and B options A and B

Total No. of machines Total No. of machines Total No. of Total No. ofh/year h/year h/year machines h/year machines

Single stem 763 1 2821 1 16678 7 248 11st coppice 978 1 3672 2 21705 8 395 12nd coppice 929 1 3483 2 20588 8 384 13rd coppice 1012 1 3793 2 22424 9 431 1

Seasonal Forwarder Feller–buncher Tractor fork—loaderharvest options C and D options E and F options C, D, E and F

Single stem 4169 6 4389 7 249 11st coppice 5426 8 5712 9 449 12nd coppice 5147 7 5418 8 321 13rd coppice 5606 8 5901 9 521 1

Fig. 2. Whole trees and coppiced stems delivered to the headland by a forwarder and then left to transpirationally dry before furtherprocessing.

Based on the anticipated crop yields and the pe-riod available for harvesting and transporting the largevolumes of biomass, the numbers of each type of ma-

chine needed to meet the demands of each systemoverall were calculated. For all-year-round harvestingwhere only small areas need to be cut perhaps every

R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37 33

3–4 weeks, simple harvesting methods are feasible.For this reason, the tractor-mounted saw followed byseparate collection by tractor and fork loader or grabwas chosen. It was assumed that the machines used inOptions A and B would be working full time on theproject all year round whereas the -eld machines usedin Options C, D, E and F would be only employed for60 days a year or less on this project and would beavailable for other tasks oK-season.For seasonal harvesting systems where a fast work

rate is necessary to harvest the total area within the60 day period, larger machines are needed and alsoat times, multiples of them (Table 2). This approachis similar to harvesting grain where a limited seasonexists and many combine harvesters have to be em-ployed within a cropping region for an intensive shortperiod. In this study, two harvest machines were cho-sen for comparison. The forwarder (options C and D)was designed for harvesting trees using a grapple andcutting head which then loads the combined loggingtrailer for extraction to the landing site. The cutting andtransport process is relatively slow but it was assumedthat a load of approximately 15 t could be transportedto the headland as a 45 m3 volume in a suitable trailerat a work rate of 6:7 h=ha. The concepts evaluatedwere based on work rates obtained by observations ofcommercially available machines undertaken by theNew Zealand Logging Industry Research Organisa-tion and recorded in a series of LIRO reports [17].The other seasonal harvesting system considered

(options E and F) used a feller–buncher which cutsand holds several stems then carries them the shortdistance (¡ 200 m) to the landing where they arestacked in a pile for drying or collection. Cutting isrelatively rapid compared with the forwarder but onlysmall loads can be carried so transport to the landingtakes longer and 7 h=ha was the assumed overall workrate. This relates to around 8 or 9 trees/min. Such com-mercial machines are well proven for single stem har-vesting and thinning, but ideally they should also beable to handle the coppice regrowth. For both the for-warder and the feller–buncher, tracked vehicles couldbe used to minimise ground damage in wetter areas,or Sotation tyres -tted to improve manouverability. Itwas assumed the capital and maintenance costs wouldbe similar in either case.It was envisaged that the tractor-mounted fork

loader would be used to grab and collect the cut stems

and then be used for transporting them over the shortdistances. Variations of the loader could also be usedfor loading trucks, stacking the stems and unloadingmaterial from the truck/trailer. A load size of 2 m3

was assumed. Although the machine designs mightvary to some extent between options depending onthe speci-c situation, for simplicity of analysis, onlyone machine cost was used. The work rate variedwith the nature of the task and whether the machinewas working in the forest or around the landing orstorage area.Transport of the material to the power plant was by

truck/trailer with a maximum gross weight of 24:5 tand volume of 105 m3. It was assumed that the truckwould always be carrying a maximum volume butthat the payload would vary with moisture content. Around trip to load the material, transport it 24:8 km tothe power plant, unload it, and then return to the -eldto collect another load was assumed to take around2:4 h.A 145 kW diesel powered, -xed knuckle boom

loader at the power plant was to be used for unload-ing the truck/trailers and also loading the tub grinder.Electric powered versions are also available. To meetthe fuel demand of the power plant, a feed rate ofover 34 t=h is needed so this capacity was allowedfor in the selection of the equipment. The tub grinderwas located close to the feed conveyor system of theplant, and used a hammer mill and screen to givesuitable comminution in terms of an acceptable parti-cle size range. Based on the size of stems anticipatedfrom 3-year-old eucalyptus, being up to 8 m long and200 mm base diameter, it was assumed that the ham-mers will need replacing every 240 h of work and theteeth every 40 h or so which has been allowed for inthe cost calculations.The main costs and characteristics of each machine

are given in Table 3, based on [14,15,18,19] as wellas on personal communications with several local ma-chinery suppliers and manufacturers. Where not avail-able, other cost and performance data were assessedbased on [20] but updated to 2001 prices. A discountedanalysis was not undertaken. Fixed costs were simplyestimated using straight-line depreciation [21] with theuseful life, salvage value and productivity estimatedfor each individual machine.General assumptions used in the cost analysis

were diesel fuel price, $NZ0.65/l; lubricants, $3.00/l;

34R.E.H.Sims,P.Venturi/B

iomass

andBioenergy

26(2004)

27–37

Table 3Machinery performance and cost assumptions as used in the analysis

System options A B AB AB ABCDEF CD EF ABCDEF ABCDEF ABCDEFMachine type saw saw tractora tractor+fork tractor+fork large feller– truck and knuckle stationary

(high capacity) (low capacity) -eld transport loader forwarder buncher trailer boom chipper

Capital cost $ 12000 12000 80000 80000 80000 340000 200000 150000 255000 400000Salvage value $ 3000 3000 20000 20000 20000 85000 50000 37500 63750 100000Economic life h 4500 4500 12000 12000 12000 11000 11000 30000 13000 10000Capacity h/ha 1.2 4.5 a 26.6 6.65 7.0Productivity odt/h 11.0 3.0 a 0.51 66.5 2.0 1.9 12.25 34.25 34.25Scheduled work hours h/year 3000 3000 2000 3000 3000 2000 1000 2000 1820 1000Depreciation $/year 6000 6000 10000 15000 15000 51000 30000 7500 27321 37500Average yearly investment $/year 7500 7500 50000 50000 50000 212500 125000 93750 159375 250000Fixed costs $/year 6825 6825 15500 20500 20500 74375 43750 17813 44853 65000

Operating costsFuel L/h 10.00 10.00 b 8.00 7.00 10.00 8.00 20.00 23.00 15.00Oil L/h 0.38 0.38 0.60 1.00 1.00 1.00 0.60 2.00 3.00 0.60R& M $/h 0.10 0.10 15.00 9.90 15.50 15.50 9.90 15.00 15.00 15.00Saw blades $/year 1425 1425

Total operating costs $/year 77310 77310 15500 107010 121860 20392 34470 103140 44853 65000Total annual cost $/year 84135 84135 31000 127510 142360 94767 78220 120953 89705 130000

Hourly cost $/h 28.05 28.05 15.50 42.50 47.45 47.38 78.22 60.48 49.29 130.00

Cost per oven dry tonne $/odt 2.55 9.47 1.41 83.34 0.71 23.69 41.17 4.94 1.44 3.80

aTractor hourly costs are included with the saw or fork costs.bTractor fuel cost is included in the cost calculations for each machine powered by the tractor.

R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37 35

machine salvage cost, 25% of capital cost; interest at9.5%; insurance at 1.5% of capital cost; and labourat an average of $17.57/h including holiday pay,accident compensation, etc.

3. Results and discussion

A summary of the harvest, transport and processingcosts for each of the six system options compared isgiven in Fig. 3.The slightly higher biomass purchase costs for Op-

tions C, D, E and F are due to the greater volumeneeded to compensate for the dry matter losses whenin longer-term storage. In these four options, the costof transporting to the landing is included in the har-vesting costs as it is all carried out in one operation,whereas for options A and B the mechanised cuttingby circular saw is separated from the collection bytractor and fork loader.Transport costs over the 24:8 km distance to the

power plant partly depend on the moisture content ofthe carted biomass. The storage costs are less for op-tions A and B than for the seasonal system options asthe average storage time is relatively short. Chippingcosts at the plant were the same for all options.Overall the all-year-round harvesting options were

$37.90/t dry matter delivered for A and $49.10 forB. These showed slight economic advantages overthe seasonal harvest operations (which ranged from$50.21 to $72.98/t dry matter delivered), partly asa result of the reduction in dry matter losses due

0 10 20 30 40 50 60 70 80

Option A

Option B

Option C

Option D

Option E

Option F

$/t dry matter delivered to the power plant

Purchase

Harvesting

Transport to landing

Transport to power plant

Storage

Chipping

Fig. 3. A comparison of the delivered cost of biomass to theconversion plant for all-year-round harvesting options A and B,and for seasonal harvesting options C and D based on using aforwarder, and E and F using a feller–buncher.

Option A.

0

10

20

30

40

50

60

-50 0 50 100 200

Percentage change from base case (%)

$/t

dry

mat

ter

del

iver

ed

Purchase price

Harvest cost

Transport distance

Storage losses

Drying rates

Option C

0

10

20

30

40

50

60

70

80

90

-50 0 50 100 200

Percentage change from base case (%)

$/t

dry

mat

ter

del

iver

edPurchase price

Harvest cost

Transport distance

Storage losses

Drying rates

Fig. 4. Sensitivity analysis of purchase price, harvesting costs,transport distance from the -eld to the power plant, drying rateand dry matter (DM) losses for all year round harvest option Aand seasonal harvest option C.

to the shorter storage period, and partly due to thecheaper harvesting costs due to the use of smaller scalemachinery. Option A gave the lowest-delivered costswhich was equal to $1.98/GJ but this result was de-pendent on the assumption that the tractor-poweredsaw could cut the trees at the high assumed work rateof 1:2 h=ha which is yet to be proven in practice. Op-tions C–F ranged from $2.61 to $3.80/GJ delivered.

3.1. Sensitivity analysis

A sensitivity analysis was conducted to assess theresponse of the delivered costs of the biomass tochanges in purchase price, direct harvesting costs,transport distance to the power plant, and the dryingrate which leads to variable moisture contents and drymatter losses (Fig. 4). Options A and C were selectedfor this analysis being the best of each harvesting sys-tem (all-year-round or seasonal) judged by the lowest

36 R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37

$/t dry matter delivered. The base case identi-ed foreach parameter as shown in Fig. 3 was varied by−50%, +50%, +100%, and for the transport distanceup to +200%, thereby giving ranges of

• $5–$20/t purchase price;• $1.88–$7.92/t for harvesting using the saw inoption A;

• $11.85–$47.38 for harvesting using the forwarderin option C;

• 12–72 km road travel distance to the plant;• 49.6–18.4% moisture content at combustion foroption A;

• 47.6–14.0% moisture content at combustion foroption C; and

• 0.75–3% dry matter losses per month.

For option A, using a relatively cheap set of equip-ment operating throughout the year the purchase priceof the biomass had the greatest eKect on the deliveredcost. Storage losses were of little signi-cance sincethere is a shorter average storage time of approxi-mately 1 month. This resulted in the moisture contentof the deliveredmaterial being slightly higher at 39.2%wet basis for the base case, compared with option Cwhich after 6 months average storage was 35.2%. Thisis as expected since most of the transpirational dryingoccurs in the -rst few weeks after cutting.Seasonal harvesting systems as in option C requires

relatively expensive machines. Multiples of them(Table 2) are necessary to cut and handle the largevolumes of biomass to meet the harvest deadlines inthe given 8–10 week period in order to keep the plantoperating for 12 months. It is, therefore, not surpris-ing that the direct harvesting costs have the largestimpact on the delivered fuel costs (Fig. 4).

4. Conclusions

Based on the yields recorded from -eld trials ofcoppiced eucalyptus in the Manawatu region of NewZealand over four rotations, the land area to supply a10 MWe base load power station was calculated to bearound 2800 ha. Harvesting this area in a short sea-son by using expensive high-performance machineryand storing the biomass for an average of 6 monthswas shown to be a more expensive system than fre-quent harvesting of small areas throughout the year.

The most expensive seasonal harvesting option (F) at$3.80/GJ of delivered biomass was almost double thecost of the better of the two all-year-round systemsevaluated. However, several key assumptions regard-ing the use of the simple tractor-mounted circular sawfollowed by subsequent collection of the drying treesusing a tractor and fork loader, need to be con-rmedby -eld evaluation. If veri-ed, then all-year-roundharvesting of short rotation coppice eucalyptus in atemperate climate appears to be a feasible methodof signi-cantly reducing the delivered costs ofthe biomass fuel to a power plant located withina 20–70 km radius.

References

[1] Sims REH. The brilliance of bioenergy—in business andin practice. London: James and James, 2002, 320p. ISBN1-902916-28-X.

[2] Sims REH, Handford P, Bell T. Wood fuel supplyand utilization from short rotation energy plantations.New Zealand Ministry of Energy Contract 881, 1990.

[3] Hall P, Gigler JK, Sims REH. Delivery systems of forestarisings for energy production in New Zealand. Biomass andBioenergy 2001;21(6):391–9.

[4] Kofman P, Spinelli R. Storage and handling of willowfrom short rotation coppice. Elsamproject, 1997; ISBN87-986376-2-2. p. 118.

[5] Hall P. EKect of in-forest drying on cost of transportingforest arisings as bio-energy. Logging Industry ResearchOrganisation technical note TN 55. Forest Research, Rotorua,New Zealand. July 2000.

[6] Sims REH, Lowe HT, Maiava T. All year round harvesting ofshort rotation coppice eucalyptus. In: Chartier P, BeenackersAACM, Grassi G, editors. Proceedings, “Biomass for Energy,Environment, Agriculture and Industry”, eighth EuropeanBiomass Conference, vol. 1. Vienna: Pergamon Press, 1994.p. 507–14.

[7] Lowe HT. Utilization of short rotation forestry for on-sitefuel wood grown as part of an eVuent disposal scheme.Master of Applied Science thesis, Massey University Library,Palmerston North, New Zealand, 1994.

[8] Blake TJ. Coppice systems for short rotation intensiveforestry: the inSuence of cultural, seasonal and plant factors.Australian Forest Research 1983;13:279–91.

[9] Fardy P. Biomass today—a new beginning for an oldresource? R E Focus. Amsterdam: Elsevier, May/June 2002.

[10] Sims REH, Maiava TG, Bullock BT. Short rotation coppicetree species selection for woody biomass production in NewZealand. Biomass and Bioenergy 2001;20:329–35.

[11] Browne AJ, Palmer M, Hunter HA, Boyd J. Transport andsupply logistics of biomass fuels. Volume 1: supply chainoptions for biomass fuels. ETSU report B/W2/00399/REP/1.University of Westminster, London, UK, 1996.

R.E.H. Sims, P. Venturi / Biomass and Bioenergy 26 (2004) 27–37 37

[12] Sims REH, Culshaw D. Fuel mix supply reliability forbiomass--red heat and power plants. In: Kopetz H, WeberT, Palz W, Chartier P, Ferrero GL, editors. Proceedings,“Biomass for Energy and Industry”, 10th European BiomassConference. Wurzburg, Germany: C.A.R.M.E.N., 1998.p. 188–91.

[13] Lowe HT, Sims REH, Maiava T. Evaluation of a lowcost method for drying fuelwood from short rotation treecrop for small scale industry. In: Chartier P, BeenackersAACM, Grassi G, editors. Proceedings, “Biomass for Energy,Environment, Agriculture and Industry”, eighth EuropeanBiomass Conference, vol. 1. Vienna: Pergamon Press, 1994.p. 461–7.

[14] Pari L. Development of a short rotation woody crops (SRWC)harvester suitable for the temperate regions. In: O’MaraK, Jennings P, Goen Ho, Anda M, Mathew K, editors.Proceedings, World Renewable Energy Congress, February,Murdoch University, Perth, Australia, 1999. p. 39–42. ISBN0-86905-695-6.

[15] Pari L. ISMA system for mechanical harvesting of shortrotation woody crops (SRWC). In: Kyriotis S, BeenackersAACM, Helm P, Grassi A, Chiaramonti D, editors.Proceedings, “Biomass for Energy and Industry”, -rstWorld Conference, Sevilla. London: James & James, 2000.p. 1966–70.

[16] Vaughan L, Shula RG. Options for harvesting and processing-rewood. The -rewood venture: planning, execution,evaluation. Forest Research Institute bulletin No. 137 (revisededition). Rotorua, New Zealand: FRI, 1993. p. 68–79.

[17] Hall P. Collection and transportation of logging residues.A system for collecting, transporting and processing loggingresidue. LIRO Report, vol. 20, No. 16 and LIRO technicalnote TN-23. Logging Industry Research Organisation,Rotorua, New Zealand, 1995.

[18] Watters MP, Mitchell CP, Sims REH, Lowe HT. Machineryoptions for harvesting, extracting and comminuting fuelwoodfrom short rotation forestry. In: Chartier P, BeenackersAACM, Grassi G, editors. Proceedings “Biomass forEnergy, Environment, Agriculture and Industry”, eighthEuropean Conference, vol. 1. Vienna: Pergamon Press, 1994.p. 272–84.

[19] Huisman W, Venturi P, Molenaar J. Cost of supply chainsof Miscanthus giganteus. Industrial Crops and Products1997;6:353–66.

[20] Witney B. Choosing and using farm machines. Harlow,England: Longman, 1988. ISBN 0582 45600 2.

[21] Miyata ES. Determining -xed and operating costs of loggingequipment. USDA Forest Service, General technical reportNC-55, United States Department of Agriculture, Washington,1980.