The biomass, in the world, constitutes the fourth largest source of energy (after coal, oil and natural gases), contributing to 14% of the planet energy needs. Nowadays, important developments in combustion technologies are making possible the rational use of wood as source of biomass. Such technologies allow to reduce dramatically the pollutant emissions, increasing also the energy efficiency of the woody materials to values above 85%.

Large quantities of ligneous biomass can be obtained from pruning operations carried out in Mediterranean orchard plantation.

According the available literature, biomass produced in fruit plantations is rarely removed and used for bioenergy production. This is due to unsolved technical problems that occur during the harvesting operations, as well as to the lack of information concerning the quantity and the quality of residues obtainable from different plantations. However, in recent years, the waste to energy research increased. This led to draw more attention on pruning as possible source of energy. More specifically, several studies have confirmed that agricultural residues may be a suitable source of biomass for the production of electrical energy.

In Italy, about 1.184 million of hectares are planted with olive trees, and more than 32% of these are located in Puglia region. Thinking that only from Puglia are obtainable over 813,000 tons of dry matter per year.

At present these residues are usually destroyed by in field burning of crushing on the soil so there is no direct economic benefit.

Since a few years, a number of machine manufacturers have been offering dedicated implements for recovering pruning residues and turning it into a biomass product. These machines have aroused considerable interest and their performance has now been documented in several studies, especially for what concerns harvesting productivity and cost. Much less is known about harvesting losses and product contamination, especially when working on olive orchards. Losses and contamination are possibly related, and may partly depend on machine settings. Losses can be reduced by lowering the pick-up device, in order to catch the residues closest to the ground. On the other hand, working too close to the ground may increase soil product contamination with soil particles, detracting from fuel quality.

The purpose of this study was to determine harvesting losses and product contamination achieved with six commercial machines, under three different settings: standard pick-up height as conventionally set by the manufacturer; pick-up working 1 cm above ground level (lower than standard); pick-up working 3 cm above ground level (higher than standard). The null hypothesis is that there is no difference in harvesting losses and product contamination between different machines and settings.

The testing area was an olive grove of 0,45 ha characterized by a flat conformation and by a very compact sandy soil with no weeds. The plantation was 40 years old, with inter-row and intra-row spacing respectively 7.5 m and 7 m. Before the test, all trees were subjected to pruning for the annual maintenance. Several agronomic and mechanic data were taken to characterize the biomass and the operating machines.

 

Pruned material characterization

A series of surveys were carried out to determine the average length, height, and width of each windrow, as well as for estimating the pruned biomass present per hectare. Specifically, the characterization was carried out selecting two plots chosen randomly from each windrow, each one extended one linear meter. The biomass present in each plot was removed and weighed. Considering a total number of 24 windrows, 48 plots (48 linear meters) were finally analyzed.

In addition, the plotted material was subjected to other measurements in order to determine the real length of the individual branches, and their diameter. Such operations allowed to verify the homogeneity of the pruning residues among windrows, making sure that during the test each machine was put at the same experimental condition.

To determine the moisture variations three samples of both fresh and pruned material, 250 g each, were processed according the role UNI EN 14774-1: 2009.

Another operation was carried out to determine the ash content from fresh pruned material. Three windrowed samples were removed from the olive grove to obtain a control-value to be used afterwards during the measurement of the ash content in chipped materials. The ash analysis was made to evaluate the presence of non-wood impurities (soil, and/or other inert materials) into chips.

 

Experimental design and testing machines

For the experimentation, 24 windrows were created along the 12 inter-rows of the olive grove (2 windrows per inter-row). These were extended for 101,4 m in length, 0,62 m in height, and 1,52 m in width. In picture it is shown a detail of the windrows.

The experimentation was divided in three theses (A,B,C), each one designed to study the performance of each machine in different conditions. Basically, the machines were tested varying the distance between the pickup organ (mechanical part responsible of the pruning collection) and the ground to determine their efficiency at different working height. According to the working height, the machines have produced chips of different forms, presenting a variable quantity of impurities.

In the development of the thesis A the pickup were set at the minimum workable distance from the ground (1 cm). In thesis B the pickup of each machine was set at the optimal working height as suggested by the manufacturer, while in thesis C these were set at the distance of 3 cm.

During the experimental tests, each machine worked on 4 windrows, 2 for thesis A and the other 2 for the theses B and C (1 windrow per thesis). Considering the number of machine tested (six), 24 windrows were finally analyzed. The test was carried out by six commercial machines (see table below), which today represent the main Italian commercially available means for harvesting pruning residues.

 

Model	Widht (mm)	Lenght (mm)	Hight (mm)	Weight (Kg)	Trailer (m3)	Unload height (mm)
Futura 160	1,750	3,300	1,700	1,300	1.7	2,200
TSB 1900	2,080	1,900	1,600	1,600	2.8	2,200
TRP-CV 145	1,900	1,400	3,400	1,000	Unload on trailer	=
TR-RAC	2,250	1,800	1,350	1,150	1.8	2,500
Comby TR 200	2,230	4,870	2,000	2,200	5.0	2,600
Picker Kargo 200	2,300	5,300	1,960	3,500	8.3	3,300

 

 

Characterization of the chipped material

After the collection phase, the 6 tested machines have unloaded the chips produced onto the ground to form six piles, which have been used in thesis A for the determination of the bulk density as indicated by UNI-EN 15103: 2009. The procedure consisted in taking five samples from each pile to obtain 5 values of bulk density that were successively averaged to get a singular value of bulk density for each machine. This parameter, expressed in kgm-3, is useful to understand the qualitative performance of the machines.

The ash content was determined according CEN/TS 14775: 2010 by taking from each machine three samples of 1,5 kg of chipped material. This time the operation was repeated by taking into account all theses. Hence, the ash content was studied on 54 samples, 18 for each thesis.

Only for thesis B, particle size distribution was determined on 4 4-L samples per machine. Each sample was weighed and the material was divided in three dimensional classes (<5 cm, among 5 cm and 10 cm, >10 cm). These dimensional classes were weighed again for defining the percent incidence of each class on total sample weight.

Data were analyzed statistically using Statview for Windows. As a first step, data distribution was plotted and visually checked for normality. When distribution was normal, differences between treatments were tested through the analysis of variance. When data distribution violated the normality assumption, attempts were made to normalize distribution through arcsine transformation. When the attempts were successful, transformed data were analyzed as above. Otherwise, the statistical significance of eventual differences between treatments was checked with the nonparametric Kruskale Wallis multiple test.

The prunings were arranged in windrows along the inter-rows and processed by machines. For a singular windrow a quantity of 0,53 t of pruned material was calculated. This corresponds approximately to 14 t/ha of fresh product or 11 t/ha of dry product. In following table are shown some data concerning the characterization of the experimental fields and the windrowed material.

 

Olive tree variety	Coratina
Planting system (m)	7.5x7
Windrow lenght (m)	101.4
Windrow hight (m)	0.62
Windrow widht (m)	1.52
Quantity of pruning per hectar (t)	14
Moisture of windrowed material (%)	22

 

Chips characterization

The chopped material generated by each machine in thesis B was divided in dimensional classes. The dimensional characterization is an important aspect to be considered; in fact, to work excellently, the power plants need materials of optimal length, which should be no longer than 10-12 cm. This because some types of devices used to transfer the woody material in the furnace (augers), could encounter problems of functionality when chips are longer than those measures. The analysis carried out has shown that all 6 machines have produced chips of acceptable dimensions. The largest part of the samples analyzed was in fact no longer, or slightly longer than 10 cm. In particular, the machines of SGARBI, FACMA, and TIERRE have produced respectively 98,5%, 93% and 90% of chips whose length was lower than 10 cm (Table below).

 

Particle size distribution and bulk density of shredded residue.

Make	% Product with length classes			Bulk density
	< 5 cm	5-10 cm	> 10 cm	Kg m-3
Tierre	60.10c	28.60d	11.30b	128a
Omat	24.25a	44.66a	31.09a	119a
Nobili	37.77b	42.14a	20.09c	174b
Sgarbi	77.50c	21c	1.50b	158b
Facma	42.88b	49.93b	7.19b	134a
Berti	19.76a	43.06a	37.18a	124a

 

 

As shown in the table, the higher value of bulk density detected was relative to the chips of the Nobili machine. Such a value could be justified because the TPR CV 145 is the only tested model equipped with a system that discharge chips, at high pressure, in a specific trailer that follows the machine along the working line during the field operations. Respect to the systems used by the other machines, which present their own collection dumpers, the high pressure system of the TPR CV 145 could determine a higher compaction of the chopped in the trailer. That would explain the higher quantity of product per m³.

Percent harvesting losses were very low, also due to the high pruning yield acting as a divider in the percent calculation. Once reported in absolute terms (t ha-1), harvesting losses were still low, but comparable to those reported in previous studies (Magagnotti et al, 2013).

Mean harvesting losses varied from 0.4 to 6% , and were significantly related to both machine type and pick-up settings. Losses were lowest for the Berti, Omat and Facma harvesters, and highest for the Nobili, Sgarbi and Tierre harvesters. Pick-up teeth number and length were both tested as additional factors in determining harvesting losses, but they did not result to have any significant effect.

 

Percent harvesting losses as a function of machine

type and pick-up height setting.

Make	Height setting
	A	B	C
Berti	0.4	0.6	0.9
Facma	1.0	1.3	1.9
Nobili	3.2	3.9	4.2
Omat	0.8	1.1	1.6
Sgarbi	4.6	5.4	6.0
Tierre	2.9	3.5	4.1

 

 

Results of the analysis of variance for the harvesting loss data.

Effect	DF	SS	%	F-Value	P-Value	Power
Setting	2	0.014	5.9	188.18	<0.0001	1.00
Machine	5	0.220	93.1	1225.12	<0.0001	1.00
Interaction	10	2.79*10-4	0.1	0.78	0.6496	0.36
Residual	54	0.002	0.8

Note: analysis conducted on arcsine-transformed data; %  incidence of the

sum of squares for the individual effect over the total Sum of Squares

 

 

Ash content ranged from an absolute minimum of 4% to a maximum of 6%. Mean values varied between 4.7% and 5.6% (Table VI). Ash content was significantly related to both machine type and pick-up settings, the latter having the strongest effect (Table VII). Raising pick-up height to 3 cm above ground level allowed reducing ash content by up to 1% point, compared  to working with the pick-up at 1 cm above ground level.

Percent ash content as a function of machine

type and pick-up height setting.

Make	Height setting
	A	B	C
Berti	5.6	5.6	4.5
Facma	5.1	5.0	4.9
Nobili	5.7	5.3	4.6
Omat	NA	4.4	NA
Sgarbi	5.2	4.8	4.9
Tierre	5.2	4.8	4.7

Note: figures obtained after back-transformation

of arcsine-transformed data for normalization purposes.

 

Results of the analysis of variance for the ash content data.

Effect	DF	SS	%	F-Value	P-Value	Power
Setting	2	3.292	43.4	26.27	<0.0001	1.00
Machine	4	0.677	8.9	2.70	0.0493	0.68
Interaction	8	1.739	22.9	3.47	0.0060	0.94
Residual	30	1.880	24.8

Note: analysis conducted on arcsine-transformed data; %  incidence of the

sum of squares for the individual effect over the total Sum of Squares.

 

 

At least part of the contamination was caused by the pick-up device of the harvesting machine, accidentally collecting soil and other inorganic materials. Contamination was effectively reduced by increasing the distance between the pick-up device and the soil surface.

From olive tree cultivations present in Italy, it is estimated that each year may be obtained several tons of pruning residues. Such a high quantity of available biomass would be suitable for several industrial applications. For instance, it could be used in co-combustion processes [13], or for the ethanol production. Many tests demonstrated that pruning residues can be collected ensuring the economic and environmental sustainability of the process for the wood-energy chain [14]. However, it would be achievable only using suitable collecting methods.

On the basis of the results obtained, the resulting product is too coarse for residential users, and is best suited for industrial conversion. Effective recovery implies minimizing product losses and product contamination. The former is mainly related to machine type, the latter to pick-up setting. Product losses are minimized by selecting appropriate equipment, offering a good match between work width and expected windrow width. Product contamination is minimized by increasing the distance between the soil surface and the machine pick-up, to avoid raking into the soil.