Harvesting Systems for Short Rotation Woody Crops

Bruce Hartsough and David Yomogida, University of California, Davis, CA

Bryce Stokes, USDA Forest Service, Auburn, AL

Acknowledgment

This study was funded in part by the Electric Power Research Institute under Work Agreement W04062-05, and by the USDA Forest Service Southern Station under Cooperative Agreement USDA-19-95-060. A copy of the full final report on the study, "Compilation of State-of-the-Art Mechanization Technologies for Short-Rotation Woody Crop Production", is available from the first author.

Abstract

We conducted a state-of-the-art survey of equipment and systems that are or might be used to harvest short-rotation woody crops. Most equipment currently in use in the US has been developed for conventional forest operations and is probably suboptimal for short-rotation conditions. The notable exception is the agricultural equipment-based harvesters recently developed in Scandinavia for small SRWC harvested as fuel. We reviewed potential means of improving harvesting systems for larger SRWC in the US.

Introduction

Various harvesting systems have been suggested for SRWC plantations for pulp production and biomass energy production. A system may include five functions:

1. felling,
2. in-stand transport (primary transport: skidding or forwarding),
3. separation of pulpable wood from residues (only for pulp production),
4. chipping or other comminution,
5. and stand-to-facility (secondary) transport.

Two or more of these may be combined together in one operation, making the system more compact, and utilizing less equipment. They may be reordered. Separation is not necessary in the production of fuel for direct-combustion, and chipping would not be included if the trees were to be burned in whole tree form.

Conventional Equipment

Systems Currently Used in the US for Harvesting SRWC

SRWC harvesting in the US has been geared towards the paper/pulp industry with by products targeted, in some cases, for biomass energy production. Essentially all harvesting is carried out with conventional forestry equipment (Stokes and McDonald 1994). The two common systems and equipment used to produce clean pulp chips are described below.

1. Feller/Buncher - Grapple Skidder- Chain Flail Delimber/Debarker - Mobile Chipper - Chip Van (Delimber/debarker residues are comminuted by tub grinder, then transported in chip vans.)
2. Feller-Buncher - Grapple Skidder - Irongate Delimber - Log Truck - Drum Debarker - Fixed Chipper (Residues from the delimber are left on site. Those from the drum debarker are hogged at the mill for fuel.)

Tricycle or articulated rubber-tired drive-to-tree feller/bunchers are by far the cheapest commercially available machines for felling and bunching trees in the 5" to 10" DBH range, to be followed by skidding, whole-tree forwarding or woods-mobile chipping. They cause more soil disturbance than other felling methods. Rubber-tired or tracked limited-area (excavator-style) feller-bunchers are more expensive than drive-to-tree machines, but they can travel in a single track, causing very little surface disturbance.

Rubber-tired grapple skidders are well-proven machines. They are obviously overbuilt for most SRWC plantations; the heavy guarding for machine protection, extreme axle or frame oscillation capabilities for rough terrain, low gearing for handling slopes, and decking blade may have little or no utility.

For pulpwood, separation of wood and residues is essential since bark, small branches and foliage are undesirable in the pulping process. (Whole-tree chips are being used by some pulp mills, but only as very minor fractions of their total furnish.) Delimbing of small trees is carried out by irongates within the stand, or chain flails at the landing. Currently, the two main types of debarkers used in the pulp industry are drum debarkers, typically used in short-wood harvesting operations and located at the pulp mill or a central processing yard, and chain flail delimber/debarkers, used at landings to which skidders deliver whole trees.

Chain flail delimber-debarkers are capable of handling small trees effectively, because multiple stems can be processed simultaneously. They are mobile and do not require much space, so they are used at landings adjacent to the harvesting units. They may (or may not, depending on tree characteristics) recover more pulpable fiber from branches and tops than some other methods. The main disadvantage is the inherently inefficient separation concept, i.e. using a blunt instrument to beat off the bark and limbs, which results in high chain costs and damage to the surface of the bole. Poplar is more easily broken than conifers, so smaller diameter chain must be used to prevent excessive breakage of tops. Debarked trees from the delimber/debarker are fed directly into the chipper. In recent years, manufacturers have produced machines known as delimber/debarker/chippers, which combine a chain flail and a disk chipper into a single machine. This eliminates the need for a second operator. Some new delimber/debarkers incorporate a third drum to increase chain-stem contact. Many operators are using multiple chains on each opening on the flail drum to improve debarking (Watson and Twaddle 1990).

Irongate delimbers, consisting of steel grids resembling stout fence gates, are commonly used in the southeast US. Grapple skidders back the tops of bunches of trees through the openings in the grate, stripping off the limbs. The gates are effective for small stems because many trees can be delimbed simultaneously.

Drum debarkers are used primarily for debarking, but several Scandinavian companies, and Proctor and Gamble in Florida, have us I drums to delimb and debark conifer tree sections or whole trees. Drums are massive, so are installed permanently at central processing yards or at mills. A few mobile models are available.

Essentially any chipper will produce chips that are acceptable for the direct combustion energy market. The ideal pulp chip, however, has relatively tight size tolerances, and larger disk chippers produce the highest quality chips. The blade and anvil on a disk chipper can be set to control chip thickness, and bigger chippers with higher inertia travel at more uniform speeds. Large chippers at fixed installations are powered by synchronous motors and turn at essentially constant speed. Knives on chippers at fixed installations are probably less susceptible to damage from rocks and can be replaced at more uniform intervals. All these factors result, in theory, in better chips from large fixed chippers than from small mobile chippers.

Drum chippers can process larger and less-uniform material than an equivalent-sized disk chipper. Knives on drum chippers are less susceptible to damage by rocks and dirt, and can be sharpened many times without removing the knives, therefore drum chippers are commonly used to produce biomass fuel.

Other Conventional Harvesting and Processing Equipment, not Currently Used for SRWC

Cut-to-Length Systems (Harvesters and Forwarders): Harvesters are much more expensive than feller/bunchers, but are used in many countries or regions where it is desirable to leave residues on site rather than accumulating them at roadside or using them for fuel. They also cause almost no site disturbance, and can create a mat of slash that is traveled on by the forwarders which transport the log lengths to roadside. They have been used to simultaneously delimb and debark eucalyptus. Disadvantages include the extra cost of processing trees into shorter lengths, and the extra downstream costs of handling the multiple smaller pieces. The use of a log-length forwarders has been shown to cause less soil compaction and disturbance than a skidder. Forwarders may cause less damage to stumps, but this is not a consideration for SRWC plantations that are replanted.

Boom/Stroke Processors and Single-Grip Processors typically handle only one stem at a time, although two or three stems can be roughly delimbed by some processors. They can be used at the stump or landing, but any of these machines is relatively expensive for very small trees.

Ring debarkers are used mostly with sawlogs and are located at the sawmills, but rings for smaller trees have recently been developed and may have potential for pulp material. They are the most energy efficient of existing debarking methods, because they use knives rather than impact to remove bark, and are likely to cause the least bole damage. They are usually located in a permanent yard because of their size and the auxiliary conveyers coupled with the debarker. Although combination ring delimber/debarkers have been developed, they were considerably more expensive and are not currently used. The main disadvantage of a ring is the single-stem processing and fixed lineal throughput rate for a given debarker. New designs for smaller stems have higher feed rates.

Golob (1986) proposed the use of front-end loaders for transporting bunches of small whole trees to roadside. This would eliminate the soil disturbance caused by dragging of trees, however it is probably not feasible for trees of 60 feet or longer, because of potential breakage.

Cable systems have been considered for use when soils are too wet to support tractive equipment, but tests at James River (with a Koller 300) showed their costs to be prohibitively high, as expected (Hartsough et al 1992). Productivity is low, and labor costs are high because members of the crew are idle during some parts of the cycle. Intermediate supports, which are time consuming and costly to rig, are needed at close spacing on flat ground.

Disadvantages/Limitations of Conventional Equipment

Conventional forestry machines are designed for rough terrain and a wide range of tree sizes and are therefore usually oversized and more rugged than required for most SRWC applications. While conventional machines could be redesigned for the easier operating conditions, major reductions in cost require non-conventional approaches because of the small size of SRWC trees.

Some conventional harvesting machines, such as feller/bunchers, process about the same number of trees or pieces per hour over their full range of size capability; others handle approximately fixed volumes, independent of tree size. Feller/bunchers are examples of the former, chip vans of the latter. Equipment such as grapple skidders or loaders handle fixed cross-sectional areas, and other, e.g. ring debarkers, have fixed linear throughput rates. For the fixed tree number handlers, cost per volume increases exponentially with decreasing tree size; for fixed area or length devices, costs increase but less dramatically.

Essentially every harvesting system includes some equipment that can be classified as either piece or area or length handling, so harvesting costs must be higher for smaller trees. This is strictly true only for equipment of the same concept and size. Smaller equipment of the same concept is cheaper to purchase and can be designed to process at faster rates; e.g. small ring debarkers have faster linear speeds than those of larger diameter. Labor costs per hour don't decrease, and the smaller tree volume more than offsets the faster piece or area or length handling rate, so, for the same concept, costs per volume still increase with decreasing tree size, even with optimally sized equipment. Since SRWC trees are smaller than the average trees being harvested in conventional forestry, costs per volume will be relatively high, if conventional harvesting equipment is used.

Special-purpose Harvesters for Small SRWC (dbh < 3" )

Trees with DBH values less than 3 inches are generally not suitable for pulp production; hence, this harvested woody material is usually used for biomass fuel. Two categories of harvesters are currently employed in Scandinavia: cut-and-chip and cut-only machines. A third, cut and forward harvester, has been tested but has significantly lower productivity than the other two categories. Recent tests have shown that harvesting costs are minimized with the cut-and-chip approach (Culshaw 1993). Machines based on agricultural harvesters have been most successful because the base machines were already proven, and the developments needed were relatively minor. Most efforts with purpose-built equipment have been abandoned, or are less successful.

Cut-and-Chip Harvesters

A. Purpose-Built Machines

The Bord na Mona consisted of a trailed unit pulled by a farm tractor and was designed to harvest willow up to about 3 inches in diameter (Curtin and Barnett, 1986). The project did not proceed beyond the initial stages.

The Gandini Bioharvester 93 consisted of a farm tractor with a felling and comminuting head fitted to the three-point hitch and a chip bin mounted on a steel frame on top of the tractor nose. Several problems were identified during testing (Culshaw 1993), and the project was terminated in 1994, in favor of more promising machines such as the Claas and Austoft (Spinelli, 1996).

The Salix Maskiner Harvester ('The Bender',) was designed to harvest twin rows of willow coppice, while attached to the three-point hitch of a reverse-drive Ford Versatile tractor. Stems are folded and compressed into a 'sausage-shaped' mass which may be handled, stored, and processed as though it were a log. After compression, stems are chipped; chips are blown into hitched trailers or tractor trailer units. Test runs in the UK indicated a productivity of about 4.0 ODT/standard hr, assuming a yield of 8 ODT/acre (Anon., 1995; a standard (std) hour includes allowances for servicing and personal breaks, but does not include downtime for unscheduled repairs).

The Texas A&M Harvester, primarily designed for harvesting mesquite, consists of a flail cutter head mounted on a John Deere forage harvester, and an auger and blower to convey the comminuted material to a towed van. 1994 tests in coppice regrowth stands indicated the harvester was capable of traveling about 2 mph and collecting most of the comminuted material. Another trial in a three-year-old sycamore plantation, where the trees averaged 15 feet tall and 4" diameter, was less successful. The cutter head severed stems, but it could not capture and comminute the material. Culshaw and Stokes (1995) believe that redesign of the cutter head may resolve this problem, but it is not likely that a flail cutter/comminuter can compete with more efficient approaches (saws and knives and blades). This harvester is still under development.

B. Agricultural Harvester-Based Machines

The Austoft sugar cane harvester cuts and billets stems, then conveys the billets to a trailer. As an alternative, 35 cubic feet of billets may be stored on the conveyer until a trailer arrives. Tests in the UK showed that the Austoft could average 0.9 acres/std hr or 9 ODT/std hr (Anon., 1994). Average travel speeds while cutting ranged from 1.3 to 2.4 miles/hr.

The Claas Jaguar was initially run with a slightly modified forage corn header, but the header frequently broke down and the feed control was not reliable (Culshaw and Stokes, 1995). Claas then developed a purpose-built header, which is now fully supported through their dealer network. The Claas blows chips into a chip forwarder which travels behind the harvester. During tests in the UK, travel speeds while cutting twin rows averaged 2.5 to 4.3 miles/hr, calculations indicated chat dhe Claas could produce 8.6 ODT/std hr, assuming a yield of 7.7 ODT/acre (Anon., 1995). Much higher production rates have been indicated for the Claas under Swedish conditions: 7000 cubic feet (on dhe order of 50 ODT) of chips per hour (Culshaw, 1993), and 2.5 acres per hour while cutting (Wiltsee and Hughes, 1995). These higher rates, which must be adjusted downwards to reflect expected delays, could reflect better terrain conditions and higher crop yields.

The John Deere 6910 Harvester is similar in design and method of harvesting to the Claas. Equipped with a Kemper corn header, it produced 6.8 ODT/std hr width stand densities of 13.1 ODT/acre (Anon., 1995), but dhe Kemper header was considered unsatisfactory for SRWC.

The New Holland 719 is a single-axle, single-wheeled trailed device with a header, powered by a towing tractor. The header may be detached and replaced by other units for different operations (Anon., 1994). The 719 with a corn header was tested for SRIC harvesting in the UK; it performed reasonably well in willow stands, but productivities were not measured.

Cut-Only Harvesters

Several continuous-travel cut-only harvester prototypes were developed or proposed in the 1980s and then abandoned because of the downturn in energy prices. These machines included the Virginia Polytechnical Institute/Department of Energy (VPI/DOE) Harvester (Curtin and Barnett 1986), which cut and crushed small stems to promote drying, the University of Hawaii Biomass Harvester (proposed but never built; Paquin et al 1989), the National Research Council of Canada (NRCC) FB2 and the NRCC Coppice Harvester (Curtin and Barnett, 1986).

Other continuous-travel cut-only harvesters are still under development. The Loughry Coppice Willow Harvester is mounted on the 3-point hitch of a 55 hp farm tractor. It cuts and bundles stems, then ejects tied bundles weighing about 70 lb behind the harvester (Curtin and Barnett, 1986). Tests in 2- to 4-year old stands of willow and poplar showed a range of productivities from 1.7 to 4.4 green tons/pmh (Ledin and Alriksson 1992). More recent tests of the latest version of the harvester, the Mk IV, had to be abandoned due to blockages in the feeding system. An average output of 0.04 acres/std hr (0.3 ODT/std hr) was calculated (Anon., 1994). After further tests, researchers concluded that the Loughry was unsatisfactory for SRWC (Anon., 1995).

The Frobbesta Harvester is a Swedish machine which severs stems, tilts them with an inclined auger, then releases them to fall horizontally onto a rear trailer platform. When the platform is full, the bunch of stems is pushed off. During trials in the UK, the harvester was found to have a productivity of 2.0 ODT/std hr (Anon., 1994).

Another self-propelled "cut-only" machine is the Empire 2000 which was built and demonstrated in Sweden. Stems are cut, conveyed and held horizontally in a collection chamber. Bunches may be transferred to a tractor and trailer or discharged at the end of the row. In tests in the UK, the machine achieved an overall average productivity of 6.7 ODT/std hr (Anon., 1995). The tests identified several inherent problems, which prevented the Empire 2000 from being classified as a functional machine at that time.

The Nicholson Harvester, designed in Great Britain, is intended for 1 to 2-year-old willow, (Anon., 1994). Stems are gathered between two gates, severed by two circular saws and transferred to a bundle chamber by pinch-belts. A stem counter automatically triggers a bundle tying and discharge sequence. In January 1994, a test run was carried out in the UK (Anon., 1994), indicating a productivity of 0.1 acres/std hr. Blockages were a significant problem.

Like the Probbesta, the Energiskogsmaskiner AB (ESM) Harvester carries bundles of sticks on a platform which may be tipped to form a stack.

None of the current"cut-only" machines (Empire 2000, ESM, Frobbesta, Loughry Coppice Harvester, and Nicholson) are supported by major manufacturers, but developers are willing to produce machines to order (Culshaw and Stokes, 1995).

Cut-and-Forward Harvester

The Brunn AIB SRICB Harvester was a large machine, which accumulated loads of stems up to 15,000 pounds. It cut stems and conveyed them to a bunk on the carrier, a Brunnett forwarder (Curtin and Barnett, 1986). It had low, productivity and was relatively expensive.

Special-purpose Harvester Larger Trees (dbh > 3")

Non-conventional harvesters for larger trees, including large SRWC, can be categorized as continuous-travel cut-and-chip, cut-only, cut/chip-and-forward, or cut/delimb/debark/chip-and forward. Conventional single-grip harvesters have been modified to cut-and-delimb/debark eucalyptus. In contrast to the current situation for smaller trees, none of this equipment is under active development, and only the Pallari Harvester and cut-and-delimb/debark harvesters are currency being manufactured.

Cut-and-Chip Harvesters

The Nicholson-Koch Mobile Harvester was primarily developed to recover biomass residues left by conventional timber harvesting systems in natural stands (Curtin and Barnett, 1986). Chips were blown into a trailing forwarder. The Harvester was quite large: 33 feet long, 9 feet wide, and 15 feet high. Tests in 1980, in a stand containing mostly residual pines averaging 6 inches DBH, indicated a production rate of approximately 1 acre/hr (20 tons/hr) (Sirois, 1981).

The Pallari Harvester cuts trees up to 4 inches in diameter with a set of rotating sickles and stationary anvils. Two vertical drums with triangular rotating inserts then direct the severed stems into a drum chipper (Curtin and Barnett, 1986). The Pallari Harvester provided the basic design for the Canadian Crab Combine; the latter, however, could harvest trees up to 8 inches in diameter. Both machines operated in a continuous motion. The Pallari is still being manufactured in limited quantities in Finland (Hakkila, 1996).

Continuous-Travel Cut-Only Harvesters (Continuous-Travel Feller/Bunchers)

The US Forest Service Harvester was a continuous-motion header for a forestry skidder. Trees were cut with a two-fluted milling cutter that could retract if the cutting rate was less than the forward speed of the harvester. After cutting the stem, the cutter sprung forward to sever the next stem. Cut stems were stored vertically in an accumulating area and then discharged as a loose bunch (Christopherson et al, 1989). No further development is planned (Thompson 1996).

The A-Line Swather was designed to harvest trees ranging from 4 to 8 inches DBH, in natural stands. Towed by a skidder, this machine consisted of a trailer with a side-mounted circular saw. When a tree was severed, a rotating "bat" struck the tree about 11 feet above the stump, directing the tree backwards into a collection bed. Simultaneously, the butt of the tree was knocked forward by a trip chain mounted behind the saw. Bunches (1 to 2 cords) were side-dumped away from the stand. In tests conducted in 1980, the A-line Swather achieved a production rate of approximately 300 trees/pmh (Curtin and Barnett,1986), considerably higher than a conventional feller-buncher working in similar stands. Production potentials, however, were not fully realized, primarily because the machine was very sensitive to stand conditions, especially stand density. The A-Line is currently in Canada but is not being used (Karsky 1996).

The Missoula Technology and Development Center (MTDC) Harvester was based on the A-Line Swather but had the collecting bunk and saw assembly mounted directly on a TimberJack 520A prime mover (Karsky, 1992). In one trial in natural stands averaging 6,000 stems/acre with most trees ranging from 3 to 8 inch DBH, the harvester's production rate was about 500 to 600 trees per hour. In September of 1990, the Harvester was tested at James River Corporation's Fiber Farm in Clatskanie, Oregon. It left stumps 3 to 4 inches high; James River desired to have trees cut at ground level to recover as much fiber as possible. During dumping, trees would occasionally remain in the bunk, decreasing the Harvester's productivity. Also, felled trees would bounce forward when they landed in the tree bunk, since the "trip chain" and "rotating bat" mechanisms did not guide the larger trees adequately. Consequently, after 3 or 4 trees were felled, the butts of the trees would begin to hang over the front edge of the bed. After several more trees were cut, the felled trees would hit the butts of the trees on the bed and fall forward rather than into the bunk (Kaiser, 1996). After this test, a "power roller" was added to the Harvester to force the felled trees past the cutting blade and back into the bunk. This addition was not field tested, and the machine is not currently being developed or used (Karsky, 1996).

The National Research Council of Canada NRCC FB7 was a continuous-travel harvester consisting of a 60 hp tractor with a cuffing/collecting/off loading header. All functions except driving and off loading were automatically sequenced by an on-board microprocessor. Also, once off loading was initiated by the operator, the remaining off loading process was automatic. As the Harvester reached a tree, sensors located in the cutting opening initiated the accumulation operation simultaneously with cutting by two 24 inch diameter, counter-rotating, inserted-tooth saws. An accumulator arm held the tree, then pushed it into a holding area After the holding area was filled with 8 to 10 trees (held vertically), the operator activated the of loading process: a grapple closed around the trees, swung them to the side, and placed them on the ground parallel to the direction of travel. The FB7 had two holding areas and off loading mechanisms, one on each side, so the operator could cut back and forth on the face of a plantation. In a three-year-old sycamore plantation (2.5 inch average dbh and 6-foot spacing), the FB7 produced 850 stems/hour (19 tons/hr) (Stokes, et al., 1986). Operational problems with hydraulic components, sensor switches, leaves and vines building up in the head, and some of the computer components, were expected to be overcome with only minor changes. Despite the very encouraging results, development of the FB7 was abandoned due to the drop in energy costs. The prototype is now at Massey University in New Zealand.

The concept for the National Research Council of Canada FB12 was similar to that of the NRCC FB7. The FB12, however, used a larger tractor as a prime mover and was a substantially larger machine, at approximately 30,000 lb. It was expected to have a productivity of about 800 trees/hr when cutting tree sizes of 10 inches to 12 inches DBH and 60 feet in height. Initial field tests on a ten-year-old hybrid poplar plantation near Kempville, Ontario and on a ten-year-old cottonwood stand near Vicksburg, Mississippi indicated that the FB12 might be a reliable and highly productive harvester. However, the FB12 was unable to efficiently hold large trees upright, because the grip on the stems was not firm enough. The prototype is currently in storage at HydMech Ltd. (the company that developed both the FB7 and FB 12 under contract to the NRCC) in Woodstock, Ontario.

Cut/Chip-and-Forward Harvester

The Georgia Pacific Biomass Harvester felled, chipped, and forwarded material while moving in a continuous motion through residual trees in natural stands. Woody material was severed by two counter-rotating cutters, and chipped material was fed into a bin towed by the harvester. This machine could comminute standing stems up to 5 inches DBH, randomly distributed across the harvester's front, at rates in excess of 13 tons/pmh (Curtin and Barnett 1986).It was never tested in SRWC plantations and is not being developed any further.

Cut/Delimb/Debark/Chip-and Forward Harvesters

Examples include the MB-Trac and the Bruks IF 300 Chipmaster Harvester. The latter, built on a forwarder chassis, was intended to cut and chip small trees from thinnings, and included delimbing knives and a flail debarker to produce clean chips for pulp. Trees up to 12 inches in diameter could be felled using a head on a long crane (Froding, 1989). A disk chipper blew chips into a 530 ft 3 chip bin, which could be unloaded by dumping sideways (Anon., 1989). The multiple functions resulted in a costly machine whose production rate was too low to be economical.

Cut-and-Delimb/Debark Harvesters

Examples of conventional single-grip harvesters that have been modified to debark eucalyptus as well at to fell, delimb and buck include the Bell SP35 Harvester, the Lako/Kato Harvester and the Waratah HTH. According to Bell, productivity of approximately 18 to 20 tons/hr has been achieved with the SP35 (Anon., 1991). In Australian tests, the Lako/Kato processed 30 to 70 trees/PMH (200 to 950 cubic feet/PMH) while removing 70-100% of the bark (Kerruish and Rawlins 1991). The Waratah design, based on the Lako/Kato, has proven to be more robust, reliable, and efficient in removing bark. In thinning of stringy-barked eucalyptus in Tasmania, the mean bark removal was 91 percent, and debarking, delimbing, and topping time averaged 0.71 minutes per tree (Kerruish and Rawlins (1991). In a June 1994 study on eucalyptus in New Zealand, the Waratah produced 31 tree-lengths/PMH (880 cubic feet; Gadd and Sowerby, 1995) with trees averaging 11 inches DBH. These machines leave all of the biomass at the felling site, which effectively eliminates the opportunity to utilize the residues for fuel.

Potential Improvements

Many studies indicate that harvesting and handling constitute one of the largest components of SRWC costs, and therefore one of the largest opportunities for improvement (e.g.Wiltsee and Hughes, 1995). Hartsough and Richter (1994) pointed out that conventional harvesting equipment has been primarily developed for forest conditions, where conditions such as rough and broken terrain with rocks, large stumps, and down logs exist. Additionally, these harvesting machines were generally developed for harvesting coniferous trees that are larger and less uniform in size than trees produced on SRWC plantations. These machines are currently being used on SRWC plantations intended for pulp production, because the SRWC market has been too small to allow for development of systems that are ideally suited to short rotation conditions.

Improvements over the conventional systems may come in a variety of areas, some incremental and some dramatic. In the former category, delimbing and debarking improvements and reduction in truck/trailer tare weights hold some potential. More radical improvements are possible by developing effective continuous-travel felling equipment, by combining functions to eliminate multiple handling, and by the development of an effective and economic process to upgrade whole-tree chips to pulp quality.

Continuous-Travel Felling

In felling, costs might be reduced by developing continuous-travel machines, similar to those proposed by Golob (1986) and prototyped by Hyd-Mech for the Bioenergy Program of the National Research Council of Canada. Effective derivatives of the Hyd-Mech FB7 or FB12 would eliminate the stop-and-go, forward-and-back travel pattern inherent to conventional feller/bunchers. Although limited studies show that feller/bunchers can be highly productive in short rotation plantations (McDonald and Stokes 1993), it is difficult to imagine a conventional machine competing with a continuous-travel machine over the long term. Impressive productivity results demonstrated by continuous-travel machines such as the Claas Jaguar for harvesting willow in Sweden support this view. Based on the FB7 results (Stones, et al., 1986) and Stokes' unpublished data on the FB12 performance, Hartsough and Richter (1994) estimated that current felling and bunching costs could possibly be reduced by 40 percent. Condnuoustravel machines would also eliminate the repetitive aspects of the operator's job; with current feller/bunchers, operators cut and bunch 150+ trees per productive hour.

The FB7 cut at a rate of 1000 trees per hour while traveling down the row (Stokes et al, 1986) but the FB12 prototype had trouble with bigger stems because of stability problems when trying to hold the large stems loosely and keep them upright; positive control of upright stems is a necessity. The Claas and similar machines are successful in part because they do not have to resist large overturning moments from the crop trees. For larger stems, it would probably be better to immediately lower the trees as soon as possible to lower the center of gravity; the trees could be held horizontally while more were accumulated, rather than holding them upright. This would reduce the size, weight and boom strength requirements for the felling equipment. An analogous situation is the comparison between feller/bunchers and feller/directors; the latter are much lighter for the same-sized tree because they do not attempt to keep the trees vertical.

Accumulated bunches could be dropped on the ground or off loaded onto trailers in the field. In the long run, continuous-travel machines of some type will be the best option, but they will involve development costs, which may be substantial.

Combining Multiple Functions

To be effective, all functions on a multifunction machine must be well-utilized. Two concepts -feller/loaders and feller/chippers -- have the most potential, because each is relatively simple and could be applied in both the pulp and biomass fuel areas. (The size of the equipment would probably be smaller for the biomass market.)

A. Feller/Loaders

Many agricultural crops such as tomatoes and sugar beets are loaded directly onto on-highway transport trailers in the field by the harvester. A similar concept for tree harvesting and transport has been proposed by several individuals, including the proponents of whole tree burners (e.g. Schaller, et al, 1993). Existing excavator-style feller/bunchers could be used as feller/loaders initially, and a continuous-travel machine eventually developed.

B. Feller/Chippers

For short-rotation (three-year) willow energy crops in Sweden, continuous-travel feller/chippers such as the Claas Jaguar are the best alternatives tested to date; they are superior to continuous travel feller/bunchers or feller/forwarders because all processing is done by one machine, and the material is handled downstream as a bulk commodity rather than individual pieces. Production rates of these machines have been very impressive in some cases, on the order of those for conventional forestry equipment with large trees, even though the willow is less than 3" DBH. Chips are blown into separate chip bins, towed by agricultural tractors. The advantages are a minimum amount of equipment and minimum handling. An effective chip/residue separation method is needed to make this concept feasible for the pulp industry.

C. Feller/Forwarders

If on-highway transport vehicles cannot be towed through the field, then continuous-travel feller./forwarders may provide reasonable alternatives. Examples include the commercial Koehring KFF and the continuous-swathing A-Line and its descendent, the MTDC Swather. (The latter two are really feller/bunchers, but with an automated cut-and-deliver-to-bunk cycle would be useful in SRWC.)

D. Less-Attractive Combinations

A few feller/delimber/debarker/chippers such as the MB-Trac and Braks IF 300 have been developed (Froding, 1989), but they did not produce at economic rates. As with delimber/debarker/chippers at the landing, the delimbing/debarking function is limiting.

Feller/chipper/chip forwarders are similar to feller/chippers, but have their own integral chip bin. They may be reasonable options for small-tract, low-production operations where move-in costs are major considerations. In large tracts or dense stands they are more expensive per ton than the two-machine combination (feller/chipper and separate chip forwarder) because the felling and chipping equipment is idle while the machine is traveling with a full bin. Examples include the modified Brunnett, L1)GSET (Hall, 1995), HAFO, Bruks and Silvatec/Hedelskebet machines developed in Scandinavia

Separation of Pulpwood from Residues

A. Delimbing/Debarking

Chain costs represent the single largest operating cost component for chain flail delimber/debarkers. Stokes and Watson (1989) estimated that chain costs represent approximately 20 to 28 percent of the total flailing costs ($0.8 to $1.6 per BDT) assuming a life of 25 PMH per set of chains. But empirical tests on hardwoods in 1989, 1991, and 1994 have shown these early estimates to be low (Hartsough and Richter, 1994); chain costs may be as much as $5/BDT of chips. This discrepancy may be due to the differences between the strength properties of bark on conifers and hardwoods. These differences might be exploited to design a more efficient debarking method for hardwoods.

Chain flail delimbing/debarking is considered the bottleneck in converting trees to wood chips at the landing. A more efficient concept such as a fast ring debarker (probably located at a central yard or mill) may be preferable in the long run.

B. Upgrading Whole-Tree Chips

The new Massahake process separates whole tree chips into clean chips and residues. It has been under development in Finland for several years, and is promising because it allows whole-tree chipping at the stump or landing, and also allows landing-to-processing facility transport of highway-legal, full-capacity chip vans. Other methods for upgrading chips have been tried over the last twenty years or so, but the Massahake process is substantially different from earlier concepts: screens remove oversized chips and fines, grinders physically separate bark from the wood, a screen and pneumatic device remove more fines, then a Simco/Ramic optical sorter separates the larger pieces of bark from the clean chips, resulting in two output streams: clean chips and hog fuel. According to Gingras (1995), an industrial plant based on this process was nearing completion and start-up in Kankaanpea, Finland, and was to provide birch chips to neighboring pulp mills and hog fuel to a district heating plant. The capital cost of this plant was estimated at 14.5 million FIM (about $US 3 million) and had a rated capacity of about 1800 ft3 / hr of loose chips .

Transportation

A. Tare Weight Reductions

Recent reductions in log trailer weights (Stuart 1993) indicate a potential for similar reductions in chip van weights, but these reductions would not be unique to SRWC transport.

B. Combined Primary and Secondary Transport

Given that highway trucks and/or trailers are loaded in-field for transporting agricultural crops, it appears feasible and highly desirable to do the same with SRWC, assuming that the trees or chips will be processed at a site other than the landing. This concept could be applied with log trailers or chip vans, used in combination with feller/loaders or feller/chippers. The trailers or vans could be towed in the field by standard on-highway tractors or by agricultural tractors. The Fast Trac in-field/on-road tractors from the U.K. (Hall, 1995) may be a higher-traction alternative to standard highway tractors.

A major concern is the feasibility of moving on-highway vehicles through the field while soils are wet. Compaction will certainly result; it can be alleviated with tillage. But it is unlikely that an unmodified on-highway vehicle can travel on wet soils. Options include central tire inflation (CTI), larger, lower-pressure tires, or load platforms that can be transferred from an in-field transporter to an on-highway vehicle. Storage buffers to supply the mills or plants during the wet season are alternatives to wet season harvesting, but debarking of stored trees is difficult, and chip quality decreases with storage time.

C. Whole tree transport

With delimbed tree lengths or shortwood from slower-growing natural stands, weight limits are almost always reached before volume limits. Log trucks have been modified to haul whole trees, tree sections with limbs, or baled material (Axelsson and Bjorheden,l991). Changes are required to prevent limbs and tops from extending beyond the legal load dimensions, to prevent small broken material from falling from the load, and to compact the load. For whole trees or tree sections, considerable experience with conifers indicates that packing efficiency decreases, so payload weights are less than for chips or delimbed logs, resulting in higher transportation costs. Similar results might be expected with SRWC, and such was the case in limited tests with tree length short rotation poplar in western Oregon (Kaiser, 1994), although weight limits are reached with delimbed poplar in Mississippi. Limited trials with transporting hardwoods from natural stands on conventional trucks in the upper Midwest resulted in payloads of 21 to 26 green tons; few problems were encountered with crowns and limbs or material falling from the whole tree loads (Schaller, et al, 1993).

Most-promising Systems

Continuous-Travel Feller/Chipper - (Chip Forwarder) - Chip Truck - Permanent Chip/Residue Separation Facility: This is potentially one of the simplest and possibly cheapest systems, certainly for the stump-to-facility equipment. Separation equipment is the biggest question mark. If it were possible to move chip vans through the field, no chip forwarder would be needed. This is less of an issue than for whole trees because chip forwarders can rapidly transfer their loads to on highway trucks.

Continuous-Travel Feller/Loader - Log Trailer - Permanent Separation/Comminution Facility: Load weights and on-highway permitting are the main issues with this system, unless off-public road hauling is used. Existing feller/bunchers could be used at present, and an optimal continuous travel machine eventually developed. Choice of an optimal separation/comminution method is a question for the pulp industry. By eliminating the separation/comminution functions, the system could deliver whole trees to an energy plant.

Summary

To date, conventional forestry equipment and methods have been employed for all operational harvesting, processing and transportation of SRWC in the U.S. for pulp production. These operations are highly mechanized, the most common utilizes feller/bunchers, grapple skidders, a chain flail delimber/debarker/chipper and chip vans. Another replaces the flail/chipper and vans with irongate delimbers, log trucks and a drum debarker. All deliver clean chips to pulp mills. Residues from the flail/chipper or drum debarker may be comminuted with a tub grinder or hammer hog and transferred to an energy production facility by van or conveyer.

Conventional forestry equipment is probably not optimal for SRWC plantations; it is used by default because it is productive and reliable. The amount of SRWC harvested has not justified the full-cycle development of specialized equipment for larger trees, i.e., grater than 3" DBH. But the conditions in SRWC plantations -- flat, obstacle-free ground, small trees of uniform size growing in straight rows, uniform road spacing (in many cases), short transportation distances to the mill (in some cases), small branches and bark characteristics which differ from those of conifers -- all suggest that SRWC harvesting, processing and transportation can be carried out in different and cheaper ways.

Equipment manufacturers and researchers have pursued numerous alternatives for harvesting smaller trees less than 3" DBH) for energy production. Some of the most recent efforts in Scandinavia have been highly successful and have nearly reached the end of the development cycle. The best machines are based on well-developed harvesters for traditional crops such as corn or sugar cane, and involve relatively minor developments, such as headers specifically designed for harvesting small diameter hardwoods. Small, closely-spaced trees are not tremendously different than sugar cane or corn, which accounts for the relatively rapid success of the development efforts with the modified agricultural harvesters. In contrast, projects involving purpose-built machines designed from the ground up for small SRWC have mostly been terminated.

For small trees, harvesting concepts may be classified as cut-and-chip by one machine, cut-only, or cut-and-forward. Cut-and-chip appears to be the best option, because the bulk chips are cheaper to handle than whole trees, and because the harvester is smaller and has less idle time than a combination harvester-forwarder. (With cut-and-chip machines, separate machines usually carry and transport the chips.)

One apparent improvement for harvesting large SRWC involves continuous-travel harvesting, to replace the stop-and-go, back-and-forth (or swing-and-return) motion of conventional feller/bunchers. The readily negotiable terrain and straight rows are amenable to continuous straight line travel, which in theory should be faster (for the same machine power) than any other alternative. (Note that essentially all agricultural harvesters and the successful machines for small SRWC all travel continuously.) Many continuous travel prototypes have been built for larger trees, most of them for natural stands. None of these has met with much success, for a variety of reasons. The most promising machines, the National Research Council of Canada FB7 and FB 12, were intended for SRWC for energy, but funding for their development was terminated in the mid-1980s due to the drop in energy prices. Renewal of efforts with these or similar concepts would benefit all producers of large SRWC, but the task is not as easy as with smaller SRWC because of the larger mass and higher center of gravity of the bigger trees.

Several machines with multiple functions are available or have been tested for larger trees in conventional forest harvesting Examples include feller/chippers, feller/chipper/forwarders, feller/delimber/barkers (called harvesters in conventional forest terminology), feller/delimber/debarkers, and feller/delimber/debarker/chipper/forwarders. Some of these are successful, many are not. Multi-function machines tend to require fewer operators, be less reliable and may not utilize the components as fully as single function equipment, but this depends on the combination. For SRWC, combinations with potential benefits include feller/loaders, feller/chippers and feller/forwarders. In addition, equipment capable of both primary (i.e. within the plantation) and secondary (on-road) transportation would eliminate unloading and reloading at roadside.

Improvements in separation of pulp and residues might include a better alternative to the inherently inefficient chain flail, and an economical means of upgrading whole-tree chips for use by pulp mills. The Massahake process being developed in Finland may prove to be the latter.

A list of "ideal" harvesting/processing/transportation systems for large SSRC might include the following two examples (and others):

1. Continuous-travel feller/chipper, combined primary/secondary chip transport, and separation of clean chips from residues
2. Continuous-travel feller/loader, combined primary/secondary transport of whole trees, delimbing/debarking, and chipping.

Both systems could be used to produce either pulp chips or, by eliminating the separation step (and chipping in the second tree system), whole-tree chips or trees for energy.

References

1. Anon., 1989. "The solution to the thinning problem: Chipmaster!", Brusk News!, Brusk Mekaniska AB, Arbra, Sweden.
2. Anon., 1991. "SP35 Harvester Sales Brochure, Bell Ltd.
3. Anon.. 1994. "First Field Evaluations of Short Rotation Coppice Harvesters", Report 11/94, Forestry Commission, The Forestry Authority, Technical Development Branch, Ae Village, Dumfries, DG1 1QB.
4. Anon.. 1995. "Second Field Trials of Short Rotation Coppice Harvesters", Report 1/95, Forestry Commission, The Forestry Authority, Technical Development Branch, Ae Village, Dumfries, DG1 1QB.
5. Axelsson, Jan, and Rolf Bjorheden, 1991. "Truck Systems for Transportation of Small Trees and Forest Residues", In: J. B. Hudson (ed), IEA/BA Task VI Activity 2, Integrated Harvesting Systems Workshop, Oregon - California, USA, June 3-8. pp. 60-87.
6. Chistopherson, Nels, Bnan Barkley, Stig Ledin, and Paul Mitchell. 1989. "Production Technology for Short Rotation Forestry", International Energy Agreement (IEA) Report 89: 1.
7. Culshaw, Damian. 1993. "Study Tour Report", In:Culshaw, Damian (comp), IEA/BA Task IX Activity 1 -- Status of SR Forestry Mechanization Worldwide, Workshop and Study Tour, Sweden -- March 24, 1993 IEA Report, ETSU, Harwell, Oxfordshire, OXl l ORA, United Kingdom. pp. 62-75.
8. Culshaw, Damian, and Bryce Stokes. 1995. "Mechanization of Short Rotation Forestry", Biomass and Bioenergy, Volume (9): 127-140.
9. Curtin, Dennis, and Paul Barnett. 1986. "Development of Forestry Harvesting Technology: Application in Short Rotation Intensive Culture (SRIC) Woody Biomass", Tennessee Valley Authority, TVA/ONRED/LER-86/7.
10. Froding, Anders. 1989. "The Development of Pulp Chip Harvesters for Small Trees", In: Bryce Stokes (ed), Proceedings of an International Symposium, International Energy Agency/Bioenergy Agreement, Task VI -- Activity 3, Auburn University, Alabama, June 5-7. pp. 57-74.
11. Gadd, Jonathan, and Tony Sowerby. 1995. "The Waratah 240 HTH -- Debarking and Logging Tree-Length Eucalyptus Regnans", New Zealand Logging Industry Research Organization Report 20(1): 1-S.
12. Gingras, J.F. 1995. "Recent Developments in Chip Cleaning and Cut-to-Length Harvesting Technologies in Finland", Forest Engineering Research Institute of Canada, Eastern Division, Internal Report: IR-1995-06-01, June 1995.
13. Golob, T.B. 1986. "Analysis of Short Rotation Forest Operations." NRCC No. 26014. Ottawa, Canada. Division of Energy, National Research Council of Canada.
14. Hakkila' Pentti. 1996. Personal Communication with David Yomogida, May 21 & 24.
15. Hall, Peter. 1995. "Harvesting and Utilization of Logging Residue, John alneaves Travel Award - 1995", New Zealand Logging Industry Research Organization Special Report No. 18.
16. Hartsough, Bruce, and Randall Richter. 1994. "Mechanization Potential for Industrial Scale Fibre and Energy Plantations", In: Bryce Stokes and Timothy McDonald (eds), EA/BA Task 9, Activity 1 International Conference, Mobile, AL, March 1-3, 1994. pp. 65-78.
17. Hartsough, B. R., B. J. Stokes, and C. Kaiser. 1992. "Short-rotation poplar: a harvesting trial", Forest Products Journal 42(10):59-64.
18. Kaiser, Charles. 1994. Personal Communication with Bruce Hartsough, March 1.
19. Kaiser, Charles. 1996. Personal Communication with David Yomogida, May 2.
20. Karsky, Richard J.. 1992. "The MTDC Tree Harvester", United States Department of Agriculture, Forest Service, Technology and Development Program, 9251-2835-MTDC, August 1992.
21. Karsky, Richard. 1996. Personal Communication with David Yomogida, April 30.
22. Kerruish, C.M., and W.H.M. Rawlins. 1991. "The Young Eucalypt Report", CSIRO Bookshop, 314 Albert Street, East Malboume, Victoria 3002. 272 p.
23. Ledin, Stig, and Agnetha Alriksson. 1992. "Handbook on How to Grow Short Rotation Forests", International Energy Agency Bioenergy Agreement Task V, Energy Forestry Production Systems Activity, ISBN 91-576-4628-7.
24. McDonald, Tim, and Bryce Stokes. 1993. "Status of Short Rotation Forestry in the USA", In: Culshaw, Damian (comp), IEA/BA Task IX Activity 1 -- Status of SR Forestry Mechanization Worldwide, Workshop and Study Tour, Sweden -- March 24, 1993, IEA Report, ETSU, Harwell, Oxfordshire, OXl l ORA, United Kingdom. pp. 21-44.
25. Paquin, D., D. Singh, and T. Liang. 1989. "Development of a Biomass Harvester", ASAE/Canadian Society of Agricultural Engineering Meeting Presentation, ASAE Paper No. 897062.
26. Schaller, Bruce J., David Ostlie, and Ronald Sunberg. 1993. "Whole Tree Energy Design, Volume 2: Program to Test Key Elements of WTE", Electric Power Research Institute, Hydroelectric Generation and Renewable Fuels Program, Generation and Storage Division, TR101564, December 1993.
27. Sirois, Donald. 1981. "A Mobile Harvester for Utilization of Weed Trees and Residues", in Proceedings of the 1981 John S. Wright Forestry Conference, Weed Control in Forest Management, Purdue University, West Lafayette, Indiana.
28. Spinelli, Raffaele, 1996. Personal Communication with David Yomogida, June 10.
29. Stokes, B.J, D.J. Frederick, and D.T. Curtin. 1986. "Field Trials of a Short-rotation Biomass Feller Buncher and Selected Harvesting Systems", Biomass 11:185-204.
30. Stokes, B.J. and T.P. McDonald (compilers). 1994. Proceedings of the International Energy Agency, Task IX, Activity 1 Symposium "Mechanization in Short Rotation, Intensive Culture Forestry"; 1994 March 1-3; Mobile, AL. Auburn, AL: US Department of Agriculture, Forest Service, Southern Forest Experiment Station. 166 p.
31. Stokes, Bryce, and William Watson. 1989. "Field Evaluation of In-Woods Flails in the Southern United States, In: Proceedings of EA/BA Task VI Activity 2 Meeting, New Orleans, LA, May 29 - June 1. pp. 99-111.
32. Stuart, W.B. 1993. (Oral presentation on advances in logging trailer design.) Presented at ASAE International Winter Meeting, Chicago, Illinois, December 14-17.
33. Thompson, Mike. 1996. Personal Communication with David Yomogida, May 2.
34. Watson, W.F., A.A. Twaddle. 1990. " An International Review of Chain Flail Delimbing/Debarking", EA/BA Task VI, Activity 2, Aberdeen University Forestry Research Paper 1990:3.
35. Wiltsee, George A., Evan E. Hughes. 1995. "Biomass Energy: Cost of Crops and Power", TR102107-Vol. 2, Electric Power Research Institute, Palo Alto, CA 94304, October.

File posted on March 17, 1998; Date Modified: February 21, 1999