Olive Pomace Oil as a Chainsaw Lubricant: First Results of Tests on Performance and Safety Aspects

Abstract

The total loss lubrication system that is typical of chainsaws is responsible for a massive dispersion in the agro-forestry environment of highly impactful pollutants, mostly of fossil origin, often well known as carcinogenic substances, which, in addition to presenting a risk to the environment, represent an important risk factor for human health, especially for chainsaw users. During its use, the chain lubricant is dispersed from the guide bar tip in the form of droplets and aerosol, or it is adsorbed on wood residues and sawdust. Then, it is subjected to drift, settles on the ground and vegetation, and can hit the operators, who, after prolonged exposures, can suffer both irritation of the respiratory tract and dermal absorption. Such a risk factor is often amplified by the widespread use of less-expensive, sometimes illegal alternatives, such as exhausted motor oils. To mitigate said negative effects, a process has been in progress for several years that is aimed at replacing conventional lubricants with synthetic or biobased oils with increasing biodegradability. As a contribution to this process, a study has been started on the possibility of using refined olive pomace oil (ROPO) as a base stock for the formulation of a totally biodegradable chainsaw lubricant. On purpose, to improve its properties of viscosity and adhesivity, such an oil was added with a biodegradable thickening agent, obtaining four formulations with different viscosity. After a lab test and a preliminary cutting test on firewood, the formulation with 2% of thickener resulted in being the best, and 3.0 g kg⁻¹ of tert-butylhydroquinone (TBHQ), a food-grade antioxidant, was then added to form the final formulation (F₂) to be compared, in the subsequent four test sessions, to a biodegradable commercial chain lubricant (S_B). The tests were carried out without changing the chainsaw setting, on different wood species, both in forest and, with the aim of increasing the repeatability of tests conditions and comparability of results, at a fixed point. The fluids’ performances were mainly evaluated based both on the operators’ opinions and on the measurements of the chain–bar temperatures and of saw chain wear related to a predefined number of cuts. As to the destiny of the fluid dispersed during cutting, the overall dispersion was assessed by considering the average working time, the consumption of chain lubricant, and the forest area cut down daily. Eventually, the amounts of inhalable and respirable dust particles as vectors of oil residues were quantified by means of personal air samplers worn by the operators and analyzed to determine any differences in the concentration of metallic elements. The test results evidenced chain temperatures that were 0.5, 4.9, and 12.5 °C higher with F₂ relating to S_B, respectively, in the cutting of trunks of fresh Pinus, Eucalyptus, and dry Pinus. They were accompanied by chain weight losses of 89.5% and 35% higher with F₂ relating to S_B, respectively, in cutting tests of Turkey oak and Poplar. Such a greater wear, however, apparently did not affect the saw chain’s cutting efficiency with F₂, since the operators declared that they did not notice any difference between the performances of the two fluids at the time of comparison. The effects of higher wear on the chain lifetime, any deriving risks for the operator’s safety, and the possibility to reduce the wear levels observed with F₂ will be explored in a further study, e.g., through different settings of the lubricating system of the chainsaw. The results of the analyses of the air-sampled dust residues that were evidenced with F₂ showed lower concentrations of respirable and inhalable particles and of some metallic elements (Al, Mg, and Ca) than those with S_B. This behavior probably depends on the different interaction between sawdust and the two fluids, which differ according to their chemical–physical characteristics (different viscosity, composition, and additives). However, it represents a positive factor in favor of the use of the ROPO-based lubricant, emphasized by the total biodegradability of its residues that are possibly contained in the dust inhaled by the operators.

1. Introduction

The negative impact on the environmental and the food chain contamination from mineral oil-based lubricants, fuels, and technical products becomes more evident every day. Several studies containing numerous evidence of the source of this contamination are available [1,2,3,4,5]. Contamination can be derived from automotive gas exhaust, oil spillage from hydraulic devices of agricultural machines, and aerosols produced by two-stroke engines in use during infield activity (air blowers, grass trimmers, branch shakers, etc.) [6,7].

One of the most impacting dispersions of pollutants in agriculture is that seen in total loss lubrication systems connected with the use of chainsaws in agro-forestry practice for green maintenance and pruning. The use of loss lubricants is needed to ensure the proper lubrication and cooling of the chains and other devices during cutting activity [8]. Modern chainsaws release the oil from the bar tip once it has traveled around the guide bar. The lubricating oil enters the environment as aerosol, as oil adsorbed on wood residues and sawdust, and directly as droplets depositing on the ground and vegetation [9]. Most petroleum-based lubricating chain oils are well-known carcinogenic substances [10,11]. In addition, prolonged exposure to oil mists due to the high speed of the chainsaw can cause irritation of the user’s respiratory tract or dermal absorption. Such risk factors to workers’ health could certainly be amplified by the widespread use of less-expensive, sometimes illegal alternatives, such as exhausted motor oils, as chain lubricants.

Based on these considerations, our research group started to study the possibilities of technical use of refined olive pomace oil (ROPO) with different targets:

• Promoting the use of vegetable oils with additives in concentration as low as possible as lubricants along the food chain to replace mineral oil-based lubricants.
• Contributing to the reduction of mineral oil contamination of food and the environment.
• Finding alternative uses for ROPO, a solvent extracted from dried olive residues (pomace) that is an involuntary by-product of the olive oil production chain.

The main production sites of crude olive pomace oil are in Spain, Italy, and Greece. It cannot be used as it is because of the strong dark color, bad smell, and high free acidity values. The refined product, generally prepared by physical refining, is a yellow oil with low acidity and a flat taste such as that of all refined vegetable oils. The refined oil cannot be marketed in pure form, but it must be mixed with a certain amount of extra virgin olive oil, according to European Regulation 2022/2104 [12].

When our research began, one of the driving forces was represented by the very low market price of ROPO in comparison with other refined liquid vegetable oils, such as sunflower, rapeseed, and soy. With the beginning of the Ukraine War in February 2022, the vegetable oil market underwent dramatic trouble and, at this moment, ROPO has a bulk price that is higher than its seed oil competitors. All considerations reported in this paper can obviously be translated to a number of natural vegetable oils with a similar unsaturation degree. The use of vegetable oils as substitutes for mineral oils is not an absolute novelty, but the interest in this field has been recently reinforced by the results reported in some papers about the study of contamination along the food chain, from the field to the packaging of food products [13,14,15].

Vegetable-based lubricants first appeared in the marketplace in Europe in the mid-1980s. The acceptance and use of these products, particularly in the European forest industry, is widespread and growing. The two main reasons for this are concerns about workers’ occupational safety and health, and environmental protection. Vegetable oils have natural properties, including good lubricity, resistance to shear, a high flash point, and a high viscosity index. These qualities give the vegetable oils chain lubricant requirements similar to petroleum-based chain oils [16,17,18]. Our research group developed a solid background with several previous research projects; the last of them was the Agroener Project, supported by the Italian Ministry of Agriculture, leading to the design and the test of hydraulic fluids based on Crambe abyssinica and Carthamus tinctorius oils, minimally refined [19,20]. In a previous paper, the preparation and evaluation of hydraulic fluid based on ROPO was also discussed.

As mentioned before, the role of a lubricating oil for chainsaw covers different aspects, ranging from the friction reduction during use to the cooling of the moving device, until the protection of the integrity of the equipment [21]. Another less evident property, but necessary for a chainsaw oil, is the adhesiveness towards metallic surfaces, as this characteristic is fundamental to keeping the chain well-oiled during use. Considering a common chainsaw, in the tip section of the bar–chain system, where the chain completes a 180° arc to return towards the machine body, the centrifugal force is maximum and causes the outwards oil dispersion, leaving the lower part of the chain without lubricant during the cutting.

Unlike other different refined vegetable oils obtained from low-acidity crude feedstocks, ROPO comes from a starting material with a high content of free fatty acids. While acidity is easily removed, the neutralized oil contains a high concentration of partial glycerides (mainly diglycerides in a 4–7% concentration), whose molecules confer high polarity and better metal affinity to this fluid. Despite this property, in some chainsaws, we observed the spatter phenomenon during use. For this reason, a humectant–viscosifier additive was added at different doses to improve the oil performance, producing four formulations. The final formula tested was based on ROPO containing 2% of thickening biodegradable additive and 3.0 g kg⁻¹ of food-grade antioxidant (tert-buthylhydroquinone, TBHQ) to preserve the olive pomace oil from the oxidative degradation during the field tests. According to a green chemistry approach, both additives were chosen for their good characteristics in terms of both environmental and health safety and technical efficiency to obtain a fully biodegradable lubricant.

This paper describes the evolution of a study articulated in five test sessions aimed at exploring the various aspects connected to the use of ROPO-based fluid for the lubrication of saw chains. Each step of the activity provided additional information and suggestions used to set up the next step and increase our knowledge of the matter. The tests concerned the assessment of the ROPO-based fluid by comparing its performance to that of a widespread conventional lubricant for chainsaws in different operative contexts, ranging from the felling of trees in the forest to tests at the fixed point, with the aim of increasing, at each step, the repeatability of the test conditions and the comparability of the results.

Parallelly, based on previous experiences of air sampling used to determine the presence of pollutants in the dust produced during agricultural operations [22] and considering the impact that the use of chainsaws may have both on forest environments and on human health, the tests of cutting were accompanied by a series of air samplings to assess the amount of inhalable and respirable dust fractions and of metallic elements contained in the dust according to the EPA methodology [23] for essential and toxic elements in order to assess the level of exposure of the operators to wood dust and to the substances it possibly contains. According to the Directives (EU) 2019/130 [24] and (EU) 2017/2398 [25], wood dust is recognized as a carcinogenic agent that is also responsible for other pathologies of the respiratory tract; the analyses of dust emissions referred to the limit value of 2 mg m⁻³ during eight hours of work [26] of hardwood dust concentration (as an inhalable fraction), as established in said directives. The gravimetric analysis of the dust residues considered inhalable and respirable fractions and the dimensional composition of the latter according to five classes. The dust concentrations were related to the detected amounts of metallic elements and to the emissions of chain lubricants to assess the relative amounts inhaled by the operators and dispersed into the environment with the assumption that the use of a highly biodegradable lubricant such as the ROPO-based fluid can help reduce the risk to both the operator’s health and the environment.

2. Materials and Methods

The assessment of the aptitude of refined olive pomace oil (ROPO)-based fluid as chain lubricant was made by comparing its performance to that of a commercially available conventional fluid in different test sessions.

2.1. Chain Lubricants at Comparison

2.1.1. ROPO-Based Biolubricant

ROPO was provided by Casa Olearia (Bari, Italy). It is a refined product that is obtained via hexane extraction of dried olive oil residues, physically refined. The food destination of such a fluid requires its mandatory mixing with extra virgin olive oil.

Analytical Methods for ROPO Evaluation

All reported results are the average values of two independent evaluations after a repeatability requirement check. All used methods are International Standards, public, easily available, and completed by precision data.

Oil Additivation with Viscosifier/Humectant and Antioxidant

Considering the possibility that some ROPO characteristics, such as viscosity, moistening capacity, and adhesiveness, had to be increased to better perform the chain lubrication, an additive with suitable characteristics was identified. It was the BIOTAC, which was provided by Brad-Chem Ltd. (Leigh, Lancashire, UK) through the Italian representative URAI (Milan, Italy). According to the technical sheet, it is an additive based on natural rubbers and is therefore biodegradable and very sticky. The main features/benefits claimed by the manufacturer are as follows:

• Highly viscous, fluid.
• All components are biodegradable (90% by OECD 301B Guideline).
• Approved by the Food and Drug Administration (Sections 177.1030 and 184.1555).
• Quickly and easily dissolves in natural vegetable oils, including rapeseed oil. It is also soluble in TMP oleate and other synthetic esters.
• Enhances biodegradable vegetable properties.

Moreover, to preserve ROPO from oxidative degradation during the functional tests, a food-grade antioxidant was added at the same time.

Chainsaw Oil Preparation

The preparation procedure was split into two steps:

• Preparation of a mother solution. To prepare 10 L of the oil-additives blend, 1 L of ROPO was heated at 80 °C in a 2-L beaker, under stirring. Due to its very high viscosity, the BIOTAC additive also had to be preheated at 80 °C in the oven before use. When both constituents reached the set temperature, a certain amount of BIOTAC additive was added to the oil, continuing the stirring and heating until all additive was dissolved, and the mother solution was ready. During the described preparation, 3.0 g kg⁻¹ of tert-butylhydroquinone antioxidant (TBHQ—Aldrich cod. 112941, Merck Life Science srl, Milan, Italy) was added to the oil using the same procedure.
• Dilution of the mother solution in the final 10-L blend. To obtain the desired additive concentration in the final fluid, the obtained mother solution was diluted ten times to obtain the final fluid for tests.

According to this procedure, four formulations were produced: the first, F₀, only consisted of ROPO and TBHQ, and the other three, i.e., F₁, F₂, and F₃, contained, respectively, 1%, 2%, and 3% BIOTAC, with the aim of studying their behaviors as lubricants for chainsaws. The analyses were carried out according to standard methods. Each used method is reported near the measurement results in Section 3.

2.1.2. Conventional Lubricant

The four ROPO-based formulations just described were compared with the oil Stihl Bioplus (S_B in the following), a very popular good-quality biodegradable chain lubricant (kinematic viscosity at 40 °C: 37.19 mm²·s⁻¹). It is a bio-oil, and, as such, it is less harmful to the environment than mineral lubricants.

2.2. Tests

Several test sessions on different chainsaws lubricated by the fluids described in Section 2.1 were carried out under different work conditions. The aim of the tests was to observe the various aspects of ROPO use and to evaluate its attitude compared to a conventional chain lubricant on the market. The comparison was based on various criteria, such as the following:

• Evaluations expressed by the operators.
• Observation of the temperature of the chain and of the guide bar during cutting, under the hypothesis that any different lubricating capacity of the tested fluids could lead to different friction values in the interface between the bar and chain and consequent to different heating levels.
• Assessment of the wear of the chain through mass measurements before and after cutting activity.
• Quantitative and qualitative analysis of the air-borne particulate sampled during cutting with the chain lubricants used.

These criteria were progressively introduced into five test sessions, whose setup varied depending on the specific aims of each of them. Therefore, each test was indicated by a “T” followed by a number (0 to 4).

During a preliminary wood trunk-cutting test (T₀), two chainsaws were used with four different fluids formulated with ROPO at four different concentrations of the thickening agent (0, 1, 2, and 3%). The second test (T₁), lasting two days, concerned the professional felling of forest trees by three specialized workers utilizing two ROPO formulations in comparison with a commercially available biodegradable oil, which lubricated two professional chainsaws. The third test (T₂) provided interesting results concerning the operability of the best formulation of the vegetable fluid tested in the previous tests, in comparison with the same conventional fluid, during cutting tests at the fixed point, utilizing two different chainsaws, one of which was operated by a wireless electric motor. The fourth and the fifth tests (T₃ and T₄) regarded the same two oils and the endothermic chainsaw used in T₂ which were used in the intensive crosscutting of trunks at the fixed point and in the felling of a 15-year-old poplar grove, respectively.

Their general layout is summarized in

Table 1. Layout of the tests.

Tests	T0	T1	T2	T3	T4
Location	Bomarzo, Italy	Mugnano, Italy	Monterotondo (Rome), Italy
Period	October 2020	February 2021	June 2021	December 2021	February 2022
Duration (days)	2	2	1	1	1
Activity	Firewood cutting	Forest cutting	Intensive cutting at a fixed point	Intensive cutting at a fixed point	Felling of a poplar grove
Operators	1 worker	3 lumberjacks	1 worker	1 worker	1 worker
Chain lubricants	F0, F1, F2, F3	F0, F2, SB	F2, SB	F2, SB	F2, SB
Trunk dimensions	Variable size	Diam.: 25–40 cm	Diam.: 15–20 cm Length: 2.5 m	Diam: 30–35 cm Length: 2.5 m	Diam: 15–20 cm Length: 12–18 m
Tree species	Turkey oak	Turkey oak	Pinus, Eucalyptus	Turkey oak	Poplar
Measurements carried out during tests	-	Diameter of cuts	Diameter of cuts	Diameter of cuts	Diameter of cuts
	-	Cuts number per full of oil	Number of cuts	Number of cuts	Number of cuts per full of oil
	Bar and chain temperature	Bar and chain temperature	Bar and chain temperature	Bar and chain temperature	-
	Oil temperature inside the reservoir	Internal temp. of freshly cut trunks	Internal temp. of freshly cut trunks	Internal temp. of freshly cut trunks	-
	Weather conditions	Weather conditions	Weather conditions	Weather conditions	-
	-	Air/dust sampling and analysis	Air/dust sampling and analysis	Chain wear (mass variation)	Chain wear (mass variation)
	-	Sampling of sawdust and analysis	Sampling of sawdust and analysis	Thermographic analysis	-
Other observations	Operator ratings

, while

Table 2. Chainsaws used in the tests.

Tests	T0		T1		T2		T3, T4
Brand	Husqvarna	Oleomac	Husqvarna	Stihl	Stihl	Pellenc 1	Stihl
Model	353	952	372 XP	362 C	MS 201TC	C21HD	MS 201TC
Engine Displacement (cm3)	51.7	51.7	70.7	59	35.2	45.0 2	35.2
Power (kW)	2.4	2.5	3.9	3.5	1.8	2.0	1.8
Max recom. engine speed (min−1)	13,000	13,000	13,500	10,000	10,500	6200	10,500
Fuel reservoir volume (L)	0.50	0.50	0.77	0.60	0.30	-	0.30
Oil reservoir volume (L)	0.28	0.27	0.40	0.32	0.22	0.25	0.22
Oil pump type	Adjustable flow
Chain pitch (mm)	8.0	9.5	9.5	9.5	9.5	6.35	9.5
Guide bar length (mm)	460	460	450	500	350	280	350
Ac. pressure, dB (A)	102	101	103	106	100 ± 2.5	85	100 ± 2.5
Ac. power, dB (A)	113	113	115	117	113 ± 2.5	100	113 ± 2.5
Vibration front/rear (mm·s−2)	3.1/3.2	5.2/5.5	3.5/4	3.5/3.5	3.5/3.1	2.5/2.5	3.5/3.1
Mass without cutting group (kg)	5.0	5.4	6.1	5.95	3.7	2.55	3.7

¹ Electric wireless chainsaw. ² Displacement of an internal combustion engine chainsaw with equivalent performance, according to the datasheet.

and

Table 3. Sensors and analytical methods used in the tests.

Type of Measurement		Parameters	T0	T1	T2	T3	T4
Atmospheric conditions		Temperature, R.H., Wind speed, Atm. pressure	Local bulletin https://www.ilmeteo.it/meteo/Bomarzo (accessed on 22 October 2020)	Portable weather station Kestrel® 4500 (R-P-R Ltd., Lymington, UK)
Measurements during cutting activity		Trunk length and diameter	-	Manual measurement
		Oil temp. (in the tank)	Salmoiraghi Digital thermometer 301H +	-	-	-	-
		Chain/bar temperature	RTD PT100 probe (Resolution: 0.1 °C; Accuracy: ±0.1%)	Infrared thermometer Mestek IR01D (Range of measurement −50 °C/800 °C; Accuracy: ±0.5 °C)		FLIR E54 thermal imaging camera (IR res. 320 × 240 pixels. Thermal sens. < 40 mK at 30 °C Accuracy: ±2 °C	-
		Internal wood temperature
		Air sampling	-	-	-	Five-stage impactor SKC Sioutas (Dorset, BSI DT11 8ST, UK) + L12 37 mm diameter, PTFE filters Merck (Darmstadt, Germany)	-
				Isokinetic personal air sampler Zambelli Egoplus-TT (Bareggio MI, Italy) + 37 mm diameter, PTFE filters Merck (Darmstadt, Germany)
Laboratory measurements	Kinematic visc. @ 40 °C		Ubbelohde Viscometers for 40 °C in thermostatic bath (ASTM D445)
	Kinematic visc. @ 100 °C		Ubbelohde Viscometers for 100 °C in thermostatic bath (ASTM D445)
	Viscosity index (V.I.)		Calculated from V40 and V100 (ASTM D2270)
	Humidity	Sawdust	Oven Bicasa—MFA (Bernareggio, MB, Italy)				-
	Gravimetric analyses	Sawdust on the chain	-	Precision balance Sartorius CP 622 (Resolution 0.01 g; Min. weight: 0.1 mg; Max. weight: 622 g)			-
		Filters	-	Analytical balance Sartorius ME5 (Resolution 0.001 mg; Max. weight: 5.1 g)			-
	Chemical analyses	Filter mineraliz.	-	Microwave Assisted Acid Digestion of Sediments, Sludges, and Oils (U.S. EPA Method 3051A)
		Metallic elements		ICP-MS, method UNI EN ISO 16967:2015 (CREA Lab.)		ICP-MS, method U.S. EPA 6020B:2014 (INNOVHUB Lab.)	-

report the main characteristics of, respectively, the used chainsaws and of the adopted sensors and analytical methods.

All cutting activities were carried out by professional operators who were provided with all personal protective equipment (PPE) required by the current laws and followed all required safety procedures.

2.2.1. Preliminary Test (T₀)

The layout of T₀ is reported in Table 1. This test aimed at collecting the first opinions of a professional operator on the intensive cutting of about 2000 kg of seasoned oak (Quercus robur L.) firewood (2019 felling) of various sizes (diameter between 5 and 60 cm), as shown in Figure 1. On purpose, the operator was provided with 2 L of each of the formulations F₀, F₁, F₂, and F₃, to refill the two chainsaws described in Table 2. The equipment belongs to two different categories of use. The most important difference lies in the engine rotation speed: the Husqvarna chainsaw works at a much higher speed and is therefore intended for professional use (e.g., forestry companies), while, given its lower speed, the Oleo-Mac chainsaw is more suitable for mainly hobby users. Samples of just-cut sawdust were taken to determine the wood’s humidity according to the mass difference between the wet and oven-dried samples. In addition to the operator’s opinions, the data on atmospheric conditions, as well as the temperature of the guide bar and of the lubricant in the tank, were also collected, as indicated in Table 3.

2.2.2. Felling in Turkey Oak Forest (T₁)

To deepen the information on the behavior of the ROPO-based chain lubricant, the formulations F₀ (without thickener) and F₂ (2% of thickener) were supplied to a company specialized in the forest cutting to be tested in real conditions of use represented by the felling of Turkey oak (Quercus cerris L.) trees and trunk limbing. The test setup and the professional chainsaws involved are described, respectively, in Table 1 and Table 2. On the first day of the test, both chainsaws were equipped with new chains. The Husqvarna 372 XP was tested with the abovementioned F₀ and F₂ ROPO-based formulations, while the chainsaw Stihl 362-C was used with the conventional fluid S_B. The test consisted of the felling (and limbing) of trees during the time interval required to use up the full volume of oil. At this time, the number of cuts per full volume of oil was counted, and each cut diameter was manually measured. Moreover, the main weather parameters and the temperature of the chain and of the freshly cut wood (inside trunk temperature) were measured by the sensors and instruments reported in Table 3. Two of the three operators forming the cutting team were equipped with isokinetic personal air samplers mounting PTFE filters with the aim of collecting the air-suspended sawdust particulate and obtaining information on the potentially inhaled amount of particulate and on its quality via a gravimetrical and chemical analysis of the residues. Table 3 reports the main characteristics of the air samplers and filters and the method adopted for the analyses. The sampling airflow was set at the value of 1.7 L·min⁻¹, which is suitable to collect the respirable fraction of particulate. Samples of sawdust were collected to determine the humidity via the difference between the wet and oven-dried mass of samples (Table 3). A third air sampler (same model) was used upwind, a few hundred meters from the cutting site, to collect particulate samples as blank. From the second day of the test, the team underwent its cutting activity by lubricating both chainsaws with F₀ until using up the supplied quantity (10 L) to achieve the opinions of the operators after a period spent in normal working conditions. Eventually, the operators were interviewed to find out the opinions derived from the use of the ROPO-based lubricant compared to conventional oil and the details of the work carried out daily (actual working time, duration, and number of oil fills in a day, average area of wood cut down daily, average duration of a chain, etc.).

2.2.3. Trunks Cutting at Fixed Point (T₂)

T₂ was thought to apply intensive workloads to the chainsaws and to improve the control of test conditions, with respect to T₁, to increase the comparability of the performances of the two chain lubricants. The general test layout is described in Table 1, while Table 2 shows the main characteristics of the two chainsaws used in this case: the Stihl MS 201TC with an internal combustion engine (ICE) and the Pellenc Selion C21HD with wireless electric motor (WEM). Due to the absence of exhaust gas emissions, the WEM chainsaw was considered helpful in the evaluation of the substances contained in the dust residues of the air samplings which came only from the cutting. In T₂, each chainsaw carried out 100 crosscuts on a log continuously, with each lubricant. The test was made using trunks of Pine (Pinus radiata L.) and Eucalyptus (Eucalyptus globulus Labill.) with similar sizes (Table 1). To refer all observed data to similar starting conditions, all cuts were executed by the same operator, replacing the chain before each new trunk cutting. Figure 2 shows the details of the test setups.

The operator was also equipped with two isokinetic personal air samplers with their associated PTFE filters. As in T₁, one sampler was used for the collection of total dust, while the second mounted a Dorr-Olivier sampling head for the collection of the respirable fraction. The filters underwent a gravimetric analysis to determine the concentration of dust in the air (Table 3). A third air sampler was used upwind, a few hundred meters from the cutting site, to collect particulate samples as blank. During trunk cutting, the weather conditions were monitored, and the temperature of the chain was measured every 10 cuts. The sensors used are shown in Table 3, together with the methods adopted for the analyses which were carried out by the CREA laboratory.

2.2.4. Cutting of Turkey Oak Trunks at Fixed Point (T₃)

All data relating to the setup of T₃ to the used chainsaw and to the sensors, instruments, and analytical methods are reported in Table 1, Table 2 and Table 3, respectively. The general purpose of T₃ was the same as T₂, i.e., to create controlled work conditions which allowed us to better compare the performance of the chain lubricants, F₂ and S_B. T₃ was focused on the application of a very severe workload consisting of the intensive crosscutting of Turkey oak trunks. The test involved one ICE chainsaw used in three repetitions of 100 cuts with each oil (total of 300 cuts), on trunks with a diameter of about 20 cm. A new chain was used with each oil. As in T₂, the operator featured two personal air samplers with PTFE filters to collect the sawdust particulate with reference to the inhalable fraction and the respirable fraction of sawdust particulate. The air-sampling activity was improved in this test by introducing a 5-stage impactor, positioned 2 m away from the cutting place, with the aim of characterizing the particulate matter produced with each lubricant. The impactor works at a 9 L·min⁻¹ flow rate and provides the number of particles in the following dimensional classes: >2.5 µm, 1.0 µm, 0.5 µm, 0.25 µm, and <0.25 µm, mostly relating to the respirable fraction. All filters underwent gravimetric and chemical analyses for the determination of the concentration of the dust and of metallic elements in the air. The temperature of the chain and of the internal wood (just cut) was monitored by using a thermal image camera to identify the most stressed section of the chain–bar system and observe the level of temperature reached. Eventually, after the test, both chains were disassembled and accurately cleaned and weighted to measure their mass variation as a wear indicator.

2.2.5. Felling of a Poplar Grove (T₄) at CREA

The test took place at the CREA experimental farm (Rome, Italy) and consisted of the felling of a 15-year-old poplar grove (Populus alba L., clone AF8) that was grown without an artificial irrigation system and fertilization. The test setup is summarized in Table 1. The aim of the test was to achieve additional information on any difference in the wear of the chains lubricated with ROPO-based fluid (F₂) and conventional fluid (S_B). The felling was carried out by means of the chainsaw reported in Table 2. Two plots were defined, and in each plot, one of the lubricants was used for comparison and a new chain was used. The assessment of chain wear was made by using the difference between the mass of the new chain and of the same chain after the felling (Table 3).

2.2.6. Chemical Analyses

All analytical methods used to characterize the refined olive pomace oil were standardized, and the complete list is reported, along with the measured property, in the tables discussed in Section 3.

The chemical analyses for the determination of the metallic elements in the air were carried out at the LASER-B laboratory of CREA according to the EPA methodology for the detection of essential and toxic elements [23]. The metal content was determined on filters used in personal samplers that were pretreated by a microwave mineralization process. An aliquot of each sample was transferred to special Teflon containers and subjected to an acid attack, using a microwave digester (model Start D, Milestone, Sorisole, BG, Italy). The determination of the metal content was carried out by ICP-MS (Inductively Coupled Plasma–Mass Spectrometer model 7700, Agilent, Santa Clara, CA, USA). The etching mixture used was composed of 6 mL of HNO₃ (65% v/v) and 3 mL of H₂O₂ (30% v/v). The calibration of the instrument was carried out using multi-element standards (standard mix, concentration 10 ppm in metal) prepared in an aqueous solution acidified with 1% of HNO₃ (65% v/v).

3. Results and Discussion

3.1. Results of Laboratory Activity

Table 4. ROPO fatty acid composition according to ISO 12966-2:2017 (methyl esters prepared according to cold method with KOH) and ISO 12966-4:2015.

Fatty Acid	%Area
Myristic	0.02
Pentadecanoic	0.01
Palmitic	11.17
Palmitoleic	0.85
Heptadecanoic	0.07
Heptadecenoic	0.11
Stearic	2.91
Oleic	72.38
Linoleic	10.71
Eicosanoic	0.47
Eicosenoic	0.34
Linolenic	0.59
Docosanoic	0.22
Erucic	<0.01
Tetracosanoic	0.10

shows the fatty acid composition of ROPO. The typical pattern of olive oil was confirmed, with the fatty acids being represented by oleic, palmitic, and linoleic acids. No changes were detected after additivation.

To assess the effects of the additivation described in Section 2.1.1 on the physical–chemical ROPO properties, lab analyses were carried out on ROPO as it is and the formulation F₂ (ROPO + 3.0 g kg⁻¹ TBHQ + 2% BIOTAC). The results of the characterization are reported in

Table 5. Main physical–chemical properties of the ROPO as it is and the formulation F₂.

Parameters	Methods	Units	ROPO	F2
Density @ 15 °C	ISO 12185:1986	kg·m−3	914.8	914.8
Acidity	ISO 660:2020	% Oleic Acid	0.16	0.18
Acid value	ISO 660:2020	mg·KOH·g−1	0.31	0.36
Peroxide value	ISO 3960:2017	mEq·O2·kg−1	11.70	15.72
Kinematic viscosity @ 40 °C	ASTM D445-06	mm2·s−1	40.12	47.05
Kinematic viscosity @ 100 °C	ASTM D445-06	mm2·s−1	8.52	10.02
Viscosity index	ASTM D2270-10	-	197	207
Copper corrosion (50 °C, 3 h)	ASTM D 130	Class	1	1
Rancimat Induction Time	ISO 6886:2016	Hours	4.95	25.18
Slip point	ISO 6321:2021	°C	−3	n.a.
Limpidity point	ISO 6321:2021	°C	0	n.a.
Pour point	ISO 3016:2019	°C	−9	−9
Sulfur	ISO 20846:2019	mg·kg−1	7.6	8.3
Phosphorous	ISO 10540-3:2002	mg·kg−1	0.3	4.5
Potassium	EPA 6010D:2014	mg·kg−1	<1	15
Tin	EPA 6010D:2014	mg·kg−1	2	4
Monoglycerides	EN 14105:2020	% m·m−1	0.25	0.25
Diglycerides	EN 14105:2020	% m·m−1	6.68	6.68
Ashes	ISO 6884:2008	% m·m−1	0.003	0.003
Conradson carbon residue	ISO 10370:2014	% m·m−1	0.26	0.26

. The listed parameters are typical of refined food oils. Considering the specific application related to this study, both the kinematic viscosity and the behavior of the oil when cold are of particular importance because they could represent a limit to the conservation and use of the oil in the winter period.

The comparison between the parameter values of ROPO as it is and F₂ shows the strong impact of the additivation on the kinematic viscosity and Rancimat Induction Time, while the impact on the acidity and phosphorous and sulfur content seems negligible. Other parameters, such as the mono- and diglyceride content and the cold properties, were not affected by the additivation. The considerable presence of monoglycerides (0.25% m·m⁻¹) and diglycerides (6.68% m·m⁻¹) confirms what is reported in the Introduction, and for this specific application, it could be advantageous thanks to the greater interaction of these molecules with metal surfaces. Also, the analysis of trace elements such as aluminum, silver, barium, boron, cadmium, calcium, chromium, iron, magnesium, manganese, molybdenum, nickel, lead, silica, sodium, titanium, vanadium, and zinc did not show changes, as all element concentrations remained under the detection limit, with the only exceptions being potassium (from <1 to 15 mg·kg⁻¹) and tin (from 2 to 4 mg·kg⁻¹). A further lab test was carried out to deepen the information on the impact of different concentrations of BIOTAC on the viscosity of the four formulations defined in Section 2.1.1. The results are shown in

Table 6. Effects of the additivation of the thickening agent BIOTAC on the viscosity and viscosity index of the mixture of ROPO plus 3.0 mg kg⁻¹ of TBHQ as oxidant.

Formulation	Composition	Kinematic Viscosity @ 40 °C (mm2·s−1)	Kinematic Viscosity @ 100 °C (mm2·s−1)	Viscosity Index
F0	ROPO+TBHQ (3 g kg−1)	40.92	8.68	198
F1	ROPO+TBHQ+BIOTAC (1%)	44.36	9.36	201
F2	ROPO+TBHQ+BIOTAC (2%)	47.05	10.02	207
F3	ROPO+TBHQ+BIOTAC (3%)	50.28	10.57	207

.

As expected, the additive has a strong effect on a 40 °C kinematic viscosity, which increased from 40.92 in F₀ to 50.28 mm²·s⁻¹ in F₃. Also, a 100 °C kinematic viscosity improves accordingly (from 8.68 to 10.57 mm²·s⁻¹), while the maximum viscosity index occurs in F₂ with 2% of BIOTAC. A further thickener addition causes a reduction in the viscosity index. Before starting with the in-field tests, an overview of the products available on the market was carried out, by means of Web research on MSDS and technical sheets, as summarized in

Table 7. Comparison between F₂ and several chainsaw lubricant oils available on the market (from the Web).

Lubricants	Density @ 15 °C (kg·m−3)	Kinematic Viscosity @ 40 °C (mm2·s−1)	Viscosity Index	Flash Point (°C)	Pour Point (°C)	Remarks
F2	915	47	207	>280	−9	Vegetable-based
ENI	884	100	99	252	−18	Mineral product
ENI Forest	885	105	100	240	−14	Mineral product
Papillon	925	42	-	>280	−3	Vegetable-based
UNISPAD	920	60	180	300	−10	Vegetable-based
VPM	920	45	n.a.	>230	5	Vegetable-based
TAMOIL Blue B.	880–890	95–110	>98	>240	−9	n.a.
MOL-LUB D.F.	875–895	100	92	260	−27	Mineral product
STIHL Bioplus	920–930	32–42	n.a.	>230	n.a.	Vegetable-based
BIOD ENG 100%	1090 (20 °C)	40–55	n.a.	350	−40	Vegetable-based

.

In general, the data indicate that mineral-based products have a kinematic viscosity that is higher than vegetable-based products. The latter have a viscosity of around 40 mm²·s⁻¹, with two exceptions at 32 and 60 mm²·s⁻¹. According to the relating technical datasheets, the listed vegetable-based fluids are not pure vegetable oils but blends with products of other origin (mainly mineral) and unidentified specific additives.

3.2. Tests of ROPO-Based Fluids in Chainsaws

3.2.1. Preliminary Test (T₀)

T₀ was performed on 24 October 2020 and 25 October 2020, under the atmospheric conditions reported in

Table 8. Atmospheric conditions during the T₀ test.

Ambient Conditions	Day 1	Day 2
Temperature (°C)	17	7
R.H. (%)	68	88
Wind speed (km·h−1)	25	0
Atmospheric pressure (mbar)	1018	1018

.

The first check carried out in the context of T₀ was an empirical practice commonly adopted by expert operators to visually assess the adhesion of the fluid to the chain to evaluate its fluidity and the need for flow rate adjustment by acting on the lubrication pump. It consisted of pointing the chainsaw bar towards a clean section of the trunk and operating the machine at maximum rotation speed, observing how the oil sprayed by centrifugal force was distributed on the surface. The check was repeated for the four ROPO-based formulations. When observing Figure 3, no significant differences can be observed in the appearance of the four trunk sections “sprayed” with ROPO added at increasingly thickener concentrations.

Then, the ROPO-based formulations, F₀, F₁, F₂, and F₃, were tested in both chainsaws, which are already described in Table 2, in the intensive cutting of the oak firewood, whose average humidity, determined on the sawdust samples taken during the test, resulted in being 37.6%. During the cutting with each fluid–chainsaw combination, at an interval of 30 cuts, the operator carried out measurements of the temperature at the center of the guide bar (by contact) and of the lubricant within the oil tank (by immersion). The averages of said measurements are reported in Figure 4.

The first information provided is that the temperatures observed on the second day, at 08.00 h in the morning, with an ambient temperature of 7 °C, were higher than on the first day, at 17 h in the afternoon, with an ambient temperature of 17 °C. According to the operator, this behavior was determined by the fact that most of the activity took place in the morning of the second day, which caused the equipment to heat up more regardless of the outside temperature. However, it may have been emphasized by the lower relative humidity and higher wind speed observed in the afternoon of the first day (Table 8) which favored heat dissipation despite the higher ambient temperature. As for the cutting test, Figure 4 shows that the bar temperature never exceeded 60 °C. The chainsaw Husqvarna 353 worked at higher temperatures than the Oleo-Mac 952 both at the bar–chain level and in the oil tank, likely due to the considerably higher engine speed in the Husqvarna 353. Moreover, on both test days, higher chain temperatures occurred with F₂. It is difficult to draw conclusions from these data on the different behaviors of the four formulations, considering the heterogeneous dimensions of the material to be cut and the consequent variability of the test conditions that may have heavily influenced the temperature of the guide bar. However, at the end of the test, the ratings of the operator were largely positive, as he said that he did not notice any substantial difference among the four ROPO-based formulations and between these and the conventional chain lubricants. These positive opinions also concerned the oil’s behavior at low temperatures and the absence of oil-sawdust residues and/or incrustations on the moving parts after a prolonged stop of the machines, verified by leaving both chainsaws unused for about a month, at the end of which, they restarted without any inconvenient. The positive results of T₀ encouraged the execution of the study but also evidenced, above all, the need to carry out the tests in such a way that makes it possible to reliably compare the performance of ROPO-based fluids to that of conventional chain lubricants. Furthermore, in addition to the technical performance of the fluids, the opportunity emerged to investigate other aspects of their use, such as the quantity and quality of the sawdust dispersed in the environment and its inhalable fraction, to evaluate any exposure risks for the operators. Considering the results of T₀ and that, in the preliminary test, F₂ and F₃ showed the highest I.V. (207), which indicates greater resistance to thermal stress, it was decided to continue the study using only the F₂, whose viscosity, is lower than that of F₃ (Table 6), involves less dissipation of energy in the sliding of the chain along the bar and is nearer to the viscosity of S_B.

3.2.2. Felling in Turkey Oak Forest (T₁)

Table 9. Data collected on the first day of T₁ concerning the cutting.

Repetitions	Parameters	Rep. 1	Rep. 2	Rep. 3	Rep. 4	Rep. 5	Average
Chain Lubricant		F2	F2	F0	SB	SB
Chainsaw		Husqvarna 372 XP			Stihl 362 C
Values per full of oil	Cutting time (s)	1860	1740	1500	1161	1500	1552.20
	Cuts number (N)	414	242	300	238	414	321.60
	Time per cut (s·cut−1)	4.5	7.2	5.0	4.9	3.6	5.04
	Cut frequency (N·min−1)	13.4	8.3	12.0	12.3	16.6	12.51
Values per dm3	Cutting time (s·dm−3)	4650	4350	3750	4644	6000	4679
	Cuts number (N·dm−3)	1035	605	750	952	1656	1000
Cut diameter	Average (cm)	10.10	11.50	9.97	18.53	10.80	12.18
	Max (cm)	28.00	50.00	35.00	60.00	19.00	38.40
	Min (cm)	5.00	5.00	5.00	5.00	5.00	5.00
	St. dev. (cm)	4.31	8.68	6.82	8.38	5.63	6.77
	CV (%)	42.69	75.49	68.43	45.20	52.16	56.79
	St. error	0.21	0.56	0.39	0.54	0.33	0.41
Sum of diameters per full of oil (cm)		4087	2784	2990	4410	3240	3502
Sum of diameters per liter of oil (cm·dm−3)		10,218	6960	7475	17,640	12,960	11,051
Cut section	Average (cm2)	94.71	162.93	114.43	324.52	116.45	162.61
	Max (cm2)	615.75	1963.50	962.11	2827.43	283.53	1330.46
	Min (cm2)	19.63	19.63	19.63	19.63	19.63	19.63
	St. dev. (cm2)	89.58	327.68	190.18	316.17	104.73	205.67
	CV (%)	94.57	201.12	166.19	97.43	89.94	129.85
	St. error	4.40	21.06	10.98	20.49	6.05	12.60
Sum of sections per full of oil (cm2)		38,839	39,429	34,330	77,236	34,935	44,954
Sum of sections per liter of oil (cm2·dm−3)		98,029	98,571	85,824	308,944	192,838	156,842
Chain temperature	Average (°C)	49.39	55.94	50.15	54.58	42.96	50.61
	Max (°C)	57.70	60.00	55.60	57.70	60.90	58.38
	Min (°C)	32.00	45.00	46.60	32.00	39.20	38.96
	St. dev. (°C)	10.05	2.72	4.35	6.34	7.06	6.10
	CV (%)	20.35	4.86	8.67	11.61	16.43	12.38
	St. error	0.49	0.17	0.25	0.41	0.41	0.35
Temperature of freshly cut wood (°C)		5–7

reports the main data relating to the cutting collected on the first day of the test. The measurements refer to the time required, in each replicate, to use up the full amount of oil. Considering the different capacity of the reservoir of the two chainsaws used in the test, for a better comparison, Table 9 also shows the data relating to the full use of oil and the data calculated referring to the oil unit (dm³), according to the hypothesis that the statistical indicators remained constant.

Based on the results reported in Section 3.1, F₂ was chosen for this test among the four ROPO-based formulations because of its higher viscosity index (Table 6), which theoretically should have given the oil greater stability. S_B was used in the Stihl chainsaw in two repetitions. Then, F₂ was used in the Husqvarna chainsaw by a second operator, in two repetitions. Eventually, F₂ was replaced by F₀ in the same chainsaw in a further repetition. The variability in tree size is confirmed by the very different values of the statistical indicators of cutting diameter observed among the repetitions. The average diameter varies from 9.97 cm (Rep. 3) up to 18.53 cm (Rep. 4), and the maximum diameter varies from 19.00 cm (Rep. 5) up to 60.00 cm (Rep. 4). The generally high values of standard deviation and CV testify the high variability in cutting diameter within each repetition, as well as among repetitions. The results of Mann–Whitney test (

Table 10. Results of the Mann–Whitney test for equal medians.

U\p (Same)	Rep. 1	Rep. 2	Rep. 3	Rep. 4	Rep. 5
Rep. 1	-	4.52 × 10−1	1.86 × 10−5	4.23 × 10−47	1.25 × 10−1
Rep. 2	48,425	-	1.82 × 10−3	1.59 × 10−29	6.54 × 10−1
Rep. 3	50,740	30,770	-	1.13 × 10−38	6.25 × 10−2
Rep. 4	17,009	12,071	12,650	-	1.00 × 10−4
Rep. 5	58,000	35,500	41,100	13,040	-

) on the data of diameters indicate that all repetitions significantly differ from each other regardless of the chainsaw and/or chain lubricant used.

The differences in tree size affect the presence of branches and, consequently, the proportion between felling cuts and limbing cuts and the number of cuts per liter of oil. The latter is maximum in Rep. 5 (1656 cuts·dm⁻³). Here, we observed a relatively low average cutting diameter and the minimum value of the maximum cutting diameter, which, together with a low variability (relative to the other repetitions), indicate the prevalence of limbing cuts performed with the highest cutting frequency and the maximum chain temperature (60.9 °C) observed in T₁. On the other hand, higher average temperatures of the chain were observed in Rep. 2 and Rep. 4, where the size of the trees was such as to determine larger sections of cut, with more prolonged and less frequent cuts, a sign that the width of the cutting section and the cut frequency are probably the factors that are most responsible for the thermal and mechanical stress of the chain and therefore must be considered in tests on the evaluation of the quality of chain lubricants. Figure 5 shows the results of the air sampling carried out during T₁ by means of the personal sampler set, as described in Section 2.2.2.

The total air volume sampled was normalized according to the reference temperature of 20 °C, and all determinations concerning the residues on the PTFE filters referred to the normalized volume. The results of the ICP-MS analyses on the residues on the PTFE filters show different concentrations of several metallic elements in Rep. 3, Rep. 4, and Blank. Their values seem relevant considering the relatively short sampling time (about 30 min). The main observations concern elements used in metallic alloys subject to wear. They are reported as follows:

• Boron (B), aluminum (Al), and copper (Cu): high concentrations of the three elements appear in all three filters. It can be noticed that they are unexpectedly high also in the blank filter, thus suggesting that an accumulation of these elements in the forest could have occurred during the felling activity carried out in the days preceding the test. Consequently, the quantities detected in T₁ are probably strongly affected by the previous deposits, making it difficult to discern their origins. For instance, the level of Al detected in Rep. 4 is very high compared to that of Rep. 3, where, however, it is not neglectable. It would be interesting to quantify the rate of Al coming from the previous deposits. Similar considerations can be made for B and Cu.
• Sodium (Na), chromium (Cr), manganese (Mn), iron (Fe), nickel (Ni), and barium (Ba): These elements are present in Rep. 3, sometimes in high concentrations (Na, Cr, Fe, and Ba), while in Rep. 4 and in Blank, they have low concentrations or are undetected.

Beyond the anomalous amount of Al, these data suggest the presence in Rep. 3 of a more severe wear process on the system Husqvarna 372 XP—F₀. About the origin of the above metallic elements, we must consider that both chainsaws are operated by internal combustion engines whose exhaust gas emissions could contain a relevant part of them. The analysis of the exhaust gases could help clarify this point. To this purpose, considering the engine speeds reported in Table 2, more severe solicitations should occur on the Husqvarna 372 XP chainsaw, which works at a 13,500 min⁻¹ engine speed and probably undergoes greater wear than the Stihl 362 C, whose engine works at 10,000 min⁻¹. Despite the willingness of the cutting team to collaborate in the test, its execution was somewhat conditioned, besides the great variability in the trees size, by the need not to excessively interfere with the normal felling activity, determining the impossibility to carry out the test with an inverted chainsaw–lubricant combination.

However, T₁ provided some insight into the behavior of the chain lubricants. As regards the cutting data, no relationships were found between the five repetitions, where the variability of the dimensions of the trees was the main cause of the variation in the temperature of the chain. No noticeable effects seemed to be due to the lubricants used, which, from this point of view, can be considered equivalent. On this note, the operators also declared not to have noticed any differences among F₀, F₂, and S_B concerning the cutting efficiency, the duration of the full of oil, the chain sharpening intervals, etc. The only observation concerned the excessive viscosity of all three tested fluids. The operators reduced it by adding a small amount of fuel according to a questionable, albeit widespread practice. The levels of temperature reached in all repetitions can be considered physiologic. Considering the relatively low ambient temperature recorded the night before the test, the inside wood temperatures were low as well, ranging from 5 to 7 °C (Table 9), and contributed to the heating of the chains. The temperature of the chain is affected by the combination of the cutting section and the cutting frequency. The study of the relationship between the characteristics of the sawdust emissions (inhalable fraction and contents of metallic elements) and the use of different chain lubricants is made difficult in an environment probably contaminated by previous activity and by the exhaust gases’ emissions from internal combustion engines. These indications were used to improve the study of the performance of chain lubricants by designing a test setup that is capable of reducing the typical variability of forestry work sites. Eventually, the operators of the cutting team provided some data that helped quantify the activity they normally perform. These data are shown in

Table 11. Assessment of the oil dispersion in the environment during cutting activity based on the technical characteristics of the chainsaws used in T₁ (Table 2), on the measured working times, and on the information provided by the operators involved about the daily actual working time and surface covered.

Parameters	Units Measure	Values
Volume of the chain lube reservoir	dm3	0.376
Duration of full oil	h	0.47
Actual working time	h day−1	6
Hourly oil consumption	dm3 h−1	0.80
Daily oil consumption	dm3 day−1man−1	4.80
Cutting team composition	men	3
Team daily oil consumption	dm3 day−1	14.4
Area covered by the team daily	ha day−1	0.5
Oil dispersed per unit area	dm3 ha−1	28.8
	cm3 m−2	2.88

.

Based on them, it was possible to assess that the chain lubricant dispersed in the environment per surface unit is 28.8 dm³ ha⁻¹. This value represents a heavy burden for a delicate environment such as a forest and underlines the need to proceed with the introduction of increasingly eco-sustainable products.

3.2.3. Trunks Cutting at Fixed Point (T₂)

The felling of the trees (Pinus and Eucalyptus) of T₂ was carried out by an external contractor. To improve the comparability among the different combinations of chainsaw/lubricant/wood, trunks were required to have a diameter of 15–20 cm and length of 2 m, which would allow us to make the planned 100 continuous cuts per trunk. At the moment of the test, the trunks of Pinus, despite their sufficient size homogeneity, resulted in being partly of fresh wood and partly of dry wood. Figure 6 shows the test site, the operator equipped with PPE, the two personal air samplers at breast level, and the measurement of the chain temperature. The measurement of the chain temperature was made every ten cuts until the end of each repetition, whose duration varied, as above said, from each other.

Cutting

In Figure 6, the wooden disks produced by the cutting activity are visible on the ground. Some of them were sampled to characterize the test material with reference to each of the nine repetitions. The results are shown in Figure 7, which shows the plant species the average values of wood humidity and the trunks’ diameters.

It can be noted that the presence of dry wood, with humidity values from 13% to 14% in some Pinus trunks, is confirmed, alongside trunks of fresh wood with a humidity level from 48% to 56%. The presence of dry wood combined with the intensity of cutting made it necessary to adapt the test layout to the new conditions. Figure 8 shows the layout applied. The differences between the actual layout and the planned one (shown in Figure 2) are explained by the data reported in

Table 12. Chain temperatures (°C) observed during the cutting test schematized in Figure 6 and statistical indicators. The temperature was measured every ten cuts.

Chainsaws		ICE						WEM
Chain lubricant		SB			F2			SB		F2
Repetition		Rep. 1	Rep. 6	Rep. 2	Rep. 5	Rep. 3	Rep. 4	Rep. 7	Rep. 8	Rep. 9
Wood		Pinus (fw)	Pinus (dw)	Eucal.	Pinus (fw)	Pinus (dw)	Eucal.	Pinus (dw)	Eucal.	Eucal.
Trunk diameter (cm)		20	18.5	19	20	18.5	20.5	15.3	15.3	15.8
Unit measure		°C	°C	°C	°C	°C	°C	°C	°C	°C
Number of cuts	10	39.2	63.8	49.1	44.1	67.9	56.6	64.4	44.4	42.7
	20	49.1	80.1	51.9	43.6	98.3	57.2	73.2	43.4	45.9
	30	47.1	88.8	52.6	49.1	104.2	54.4	77.6	43.3	46.9
	40	49.1	-	54.7	-	113.2	56.1	72.1	41.2	46.9
	50	48.8	-	53.7	-	112.6	52.6	-	42.2	46.8
	60	42.1	-	55.6	-	112.6	53.8	-	42.1	46.9
	70	45.9	-	54.2	-	109.7	53.6	-	-	-
	80	45.3	-	53.7	-	-	53.3	-	-	-
	90	45.2	-	52.5	-	-	-	-	-	-
	100	-	-	52.6	-	-	-	-	-	-
Statistical indicators (All data)	Min	39.2	63.8	49.1	43.6	67.9	52.6	64.4	41.2	42.7
	Max	49.1	88.8	55.6	49.1	113.2	57.2	77.6	44.4	46.9
	Mean	45.8	77.6	53.1	45.6	102.6	54.7	71.8	42.8	45.7
	St. dev.	3.4	12.7	1.8	3.0	16.3	1.7	5.5	1.2	1.6
	St. err.	1.1	7.3	0.6	1.8	6.1	0.6	2.7	0.5	0.6
	CV	7.3	16.4	3.4	6.7	15.8	3.1	7.6	2.7	3.4
Statistical indicators (30 cuts)	Min	39.2	63.8	49.1	43.6	67.9	54.4	64.4	43.3	42.7
	Max	49.1	88.8	52.6	49.1	104.2	57.2	77.6	44.4	46.9
	Mean	45.1	77.6	51.2	45.6	90.1	56.1	71.7	43.7	45.2
	St. dev.	5.2	12.7	1.9	3.0	19.5	1.5	6.7	0.6	2.2
	St. err.	1.7	5.2	0.6	1.2	7.4	0.5	2.7	0.2	0.9
	CV	11.6	16.4	3.6	6.7	21.6	2.6	9.4	1.4	4.9

.

The chain temperatures were measured during every ten cuts. The intensive cutting determined the heating of the chainsaws. The heating level seemed to be increased in the repetitions with dry wood (Pinus dw), and the cutting had to be interrupted in Rep. 3 before the planned 100 cuts. The electric chainsaw particularly suffered from overheating to the point that, several times, it stopped working due to the activation of the thermal protection device. Consequently, the planned cutting of the trunk of Pinus (Rep. 7 in Figure 2) had to be canceled. To consider the unforeseen variable fresh wood vs. dry wood, the residual parts of the fresh and dry trunks from the first four repetitions (planned) were used in two further repetitions, which were carried out after Rep. 4 (Rep. 5 and Rep. 6 in Figure 8) before starting to cut with the electric chainsaw. The residual parts of trunks used in Rep. 5 and Rep. 6 only sufficed for 30 cuts. However, they allowed us to obtain data from all combinations of chainsaw/lubricant/wood, albeit in different numbers. The data of the chain temperatures observed during T₂ are summarized in Table 12. The following indications are provided:

-

The planned 100 cuts were executed only in Reps. 1, 2, and 4 with fresh wood, whose high humidity contributed to the cooling of the chain. In all three repetitions, after 50–60 cuts, the temperature stabilized at relatively low values which, however, seem lower with Pinus (fw) (average: 45.7 °C) than with Eucalyptus (53.1 and 54.4 °C in Rep. 2 and Rep. 4, respectively).

-

In Rep. 3, with lubricant F₂, the cutting of dry wood Pinus trunks caused the raising of the chain temperature up to 113 °C, which indicated that we needed to stop the test after 70 cuts. The data for the first 30 cuts of Rep. 3 seem significantly higher than those of Rep. 6 and indicate the possibility of a minor efficiency of F₂ as lubricant under these conditions.

-

Considering Rep. 2 and Rep. 4, respectively (100 and 80 cuts on Eucalyptus), the chain temperature values resulted in being higher in Rep. 4 in the presence of the F₂ lubricant. This could be partly due to the higher starting temperature in Rep. 4 (56.6 °C) than in Rep. 2 (49.1 °C), caused by the chain overheating that occurred in Rep. 3. Slightly higher temperatures of the chain occurred, again with F₂, in Rep. 5, compared to the values of the first 30 cuts of Rep. 1 with S_B, in the cut of Pinus fresh wood.

-

Reps. 7 to 9 concerned the electric chainsaw. Its electric motor overheated and stopped working after 40 cuts in Rep. 7 (with Pinus dry wood and chain temperatures which raised up to 77 °C) and after 60 cuts in Reps. 8 and 9 (both with Eucalyptus and chain temperatures relatively low), indicating that this type of machine is not suitable for continuous cutting tasks. As regards the behavior of the two lubricants, the comparison between Rep. 8 and Rep. 9, both on Eucalyptus, shows that lower chain temperatures occurred with F₂.

The series of values of chain temperature reported in Table 12 underwent two-sample tests (Mann–Whitney) and two-sample paired tests (t-test), which always confirmed the significance of the differences in their means and medians. The additional repetitions, Rep. 5 and Rep. 6, were not considered because of the scarcity of material to be cut and of pre-weighted PTFE filters.

Air Sampling

The residues retained by PTFE filters of the two personal air samplers in the cutting operations with the two fluids were analyzed in the laboratory to determine the following: (1) the mass (mg) of particles relating to the inhalable fraction (>2.5 μm) and respirable fraction (<2.5 μm) by means of gravimetric analysis; and (2) the mass (mg) of metallic elements in the same fractions, by means of ICP-MS. Based on the sampled air volumes, all amounts are referred to by the unit of air volume and expressed as mg m⁻³. The results are shown in

Table 13. Results of the gravimetric analysis of the PTFE filter with the residues of the total inhalable dust and of the respirable dust fraction (<4 μm) relating to the air samplings carried out during the seven repetitions.

Chainsaws	Particles	ICE				WEM
Chain Lubricants		SB		F2		SB		F2
Repetitions		Rep. 1	Rep. 2	Rep. 3	Rep. 4	Rep. 7	Rep. 8	Rep. 9
Wood		Pinus (fw)	Eucal.	Pinus (dw)	Eucal.	Pinus (fw)	Eucal.	Eucal.
Wood Humidity		48.6	56.2	14.5	57.1	48	53.7	55.1
Diameter (cm)		20.25	19	18.65	20.65	15.25	15.25	15.75
Number of cuts		90	100	70	80	40	60	60
Concentration of powder as it is	Inhalable (mg m−3)	18.17	18.11	19.00	11.75	3.70	42.62	8.83
	Respirable (mg m−3)	33.99	98.50	10.51	2.92	37.95	52.45	12.90
Concentration of dust (d.s.)	Inhalable (mg m−3)	9.34	7.93	16.25	5.04	1.93	19.73	3.97
	Respirable (mg m−3)	17.47	43.14	8.99	1.25	19.73	24.28	5.79
Differences SB—F2 (dust as it is)		Rep. 1–Rep. 3		Rep. 2–Rep. 4		-	Rep. 8, Rep. 9
	Inhalable (mg m−3)	−0.83		6.36		-	33.79
	Respirable (mg m−3)	23.48		95.58		-	39.55
Differences SB—F2 (dust d.s.)	Inhalable (mg m−3)	−6.91		2.89		-	15.77
	Respirable (mg m−3)	8.48		41.89		-	18.49

.

The concentration of respirable particles shows very large variations among the repetitions, with values that seem dramatically high in Rep. 2 (98.50 mg·m⁻³) and Rep. 8 (52.45 mg·m⁻³) and suggest they can be outliers, maybe a consequence of contamination. To evaluate the percent humidity (H) of the wood that produced the dust, the concentrations of the powder as it were divided by the factor (1-H/100) calculated for each repetition, obtaining the concentration of the powder’s dry substance. The values provided by this calculation resulted in the reduced amplitude of the variations between repetitions to the point that the possible outliers could be considered reliable values. Moreover, they confirmed the same trend of the powder as it is concerning the concentration of respirable fraction vs. inhalable fraction. Table 13 also shows the differences between the concentrations of dust observed with S_B and with F₂ in the cutting of the same wood species. The differences resulted in being negative only in the cutting of Pinus wood with the ICE chainsaw. In this case, however, the cutting was carried out for S_B on fresh wood and for F₂ on dry wood, which is supposed to be capable of generating more dust. In all other cases, the differences were largely positive. The reason for this trend is not clear. It could be connected to differences in the granulometry of the sawdust induced using S_B vs. F₂, which would imply that the two fluids interact differently with the bar–chain system. This could also explain why, in the repetitions with F₂, the respirable fraction is lower (or slightly higher) than the inhalable one. Further investigations may confirm or deny this hypothesis. After the gravimetric analysis, the residues on the PTFE filters underwent a chemical analysis to determine the contents of metallic elements in the sampled air volume. Their concentration is expressed as mg·m⁻³. This analysis regarded both the respirable and inhalable dust fraction. The results are shown in Figure 9, where, to visualize the proportion between metallic elements and total dust, we also show the concentrations of the inhalable and respirable fractions calculated on the dry substance of dust, as already reported in Table 13.

The red dashed lines in Figure 9 represent the blank concentrations in the two fractions. In the inhalable fraction, the trend of the blank is anomalous probably due to a contamination of the filter. As observed for the inhalable and respirable fractions of dust (Figure 9a), the concentrations of several metallic elements resulted in being very high and, in some cases, approached the limit values. Figure 9b,c show some peaks in the concentrations of specific elements, which occur in all repetitions, albeit to a different degree. They particularly concern Na, Al, K, and Fe, and, to a lesser extent, they concern B, Mg, Ca, Cr, Ni, and Cu. As for the results of the gravimetric analysis, the metal concentrations are always higher in the respirable fraction, indicating that said elements, probably originating from the wear of components made of metal alloys, are mostly contained in smaller particles. Observing how the concentrations of the single elements vary between the repetitions, in general, it seems that, in those with F₂ (Rep. 3, Rep. 4, and Rep. 9), the concentrations are lower than in the corresponding repetitions with S_B (Rep. 1, Rep. 2, and Rep. 8, respectively), except for Al, which resulted in being higher in Rep. 4 (cutting of Eucalyptus). As regards Al, it could also be noticed that its concentration is higher mostly in the repetitions carried out on Eucalyptus, i.e., in Rep. 2 and Rep. 8, beyond that in Rep. 4.

The comparison between Figure 9b and Figure 5 shows that, in T₁ and T₂, the elements involved are substantially the same, albeit notable differences can be noticed in their concentrations, particularly concerning Na and K, which, in T₂, reach very high levels (respectively, 2 mg·m⁻³ and 1 mg·m⁻³) probably linked to the test conditions adopted, to the very high concentrations of wood dust observed, and to the different wood species used with respect to T₁. However, beyond the evaluation of the chainsaw’s cutting performance, the data confirm that the test method adopted in T₂ is also useful in the assessment of the quality of the air through the detection of the particles emitted. The method must be suitably improved as regards the uniformity of the material to be cut in order to provide more accurate and comparable data. Eventually, considering the values of the concentrations of respirable and inhalable wood dust fractions (Table 13) and of metallic elements (Figure 9), the comparison between the data of Reps. 1 to 4 with the ICE chainsaw and those of Reps. 7 to 9 with the WEM chainsaw does not show evident differences that are attributable to the presence/absence of exhaust gas emissions.

3.2.4. Cutting of Turkey Oak Trunks at Fixed Point (T₃)

Test of Cutting

The test site setup is visible in Figure 10, with the five-stage impactor and the thermal imaging camera placed near the cutting place.

The operator features the two air samplers and the PPE. Two further air samplers were placed a few hundred meters away from the test site for blank samplings of both inhalable and respirable particles. Figure 10 also shows two thermal images taken during the test.

Table 14. Main data observed during the cutting of Turkey oak trunks using the lubricants S_B and F₂.

Fluids	Rep.	Environ. Temp.	R.H.	Wood Hum.	Wood	Cuts	Cutting Time	Time	Diameter (Mean)	Section	Total	Max
					Intern.			per Cut		Surface	Cut	Chain
	Stat.				Temp.			(Mean)		(Mean)	Surface	Temp. 1
	Ind.	°C	%	%	°C	N	s	s cut−1	cm	cm2	m2	°C
SB	1	8.9	63.3	33.6	6.3	100	670	6.7	20.4	325.9	3.30	84.9
	2	11.83	59.2	35.7	8.2	100	965	9.7	23.2	424.1	4.20	122.3
	3	12.17	60.9	32.0	7.4	100	489	4.9	15.6	191.1	1.90	62.2
	Total	-	-	-	-	300	2124	-	-	-	9.40	-
	Average	10.97	61.1	33.8	7.3	-	708	7.1	19.7	313.7	3.1	89.80
	Dev. St.	1.80	2.06	1.86	0.95	-	240	2.4	3.9	117.0	1.16	30.35
	St. error	1.04	1.19	1.07	0.55	-	139	1.4	2.2	67.5	0.67	17.52
	CV	16.39	3.37	5.51	13.1	-	34	33.9	19.6	37.3	36.99	33.80
F2	4	10.1	66.3	33.1	7	100	497	5.0	15.6	191.1	1.90	78.1
	5	10.57	66.2	35.7	7.9	100	1341	13.4	26.1	535.1	5.40	104.6
	6	10.4	66.2	32.3	8.20	100	555	5.6	16.2	207.0	2.10	112.8
	Total	-	-	-	-	300	2393	-	-	-	9.40	-
	Average	10.36	66.2	33.7	7.7	-	798	8.0	19.3	311.1	3.13	98.50
	Dev. St.	0.24	0.05	1.8	0.62	-	471	4.7	5.9	194.2	1.97	18.14
	St. error	-	0.14	1.0	0.36	-	272	2.7	3.4	112.1	1.13	10.47
	CV	-	2.30	5.3	8.11	-	59	59.1	30.5	62.4	62.73	18.41

¹ Monitoring chain temperatures with the thermal imaging camera was aimed at exploring the potential of the device and was not performed systematically. The values shown are the maximum temperatures observed by the operator and are only indicative.

summarizes the main test results and the main statistical indicators.

As the test site was under a porch, it was partly protected from the wind. The humidity of the wood, measured on the samples of sawdust taken during the test, varied between 32 and 36%. The internal wood temperature was in the range of 6.3–8.2 °C. As to the dimensions of the six trunks, their length was 2.5 m, while their average diameter varied in the range of 15.6–23.2 cm. Since the test required three repetitions for F₂ and three for S_B, the trunks used with each fluid were chosen to obtain two groups (of three trunks) as homogeneous as possible. The table also reports the average section of each trunk and the total cut surface resulting from the 100 cuts of each repetition. The total cut surfaces resulted in being equal for F₂ and S_B. Conversely, the total cutting time was different for F₂ and S_B. The ANOVA on the values of the total cut sections and cutting times observed in the six repetitions showed that the differences between the two fluids were not significant for the former, while they were significant for the latter (

Table 15. ANOVA results for the values of the cutting time and average section surface to verify the presence of significant variability in such parameters between the two fluids.

Test for Equal Means
Parameters	Variability Factors	Sum of Squares	Df	Mean Square	F	P (Same)
Cutting time	Between groups	12,060.2	1	12,060.2	0.3183	0.6294
	Within groups	559,953	4	139,988	-	-
	Error	75,770.3	2	37,885.2	-	0.7503
	Between subject:	484,182	2	242,091	-	-
	Total	572,013	5	-	-	-
Average section surface	Between groups	10.67	1	11	0.0014	0.9737
	Within groups	10.28	4	25,691.2	-	-
	Error	15,390.3	2	7695.17	-	1
	Between subject:	87,374.3	2	43,687.2	-	-
	Total	102,775	5	-	-	-

).

The different behaviors of the values of the cut section surface (equal in F₂ and S_B) and of the related total cutting time (different in F₂ and S_B) can be explained by a non-linear relationship between the two parameters. With reference to the data reported in Table 14, the diagram of Figure 11 shows that increasing the trunk section causes a more than proportional increase in the cutting time.

The six points are interpolated by an exponential function with a high R². Increasing the trunk section determines the augmented surface of contact between the chain and wood, resulting in increased friction, a higher engine load, and a lower engine speed. This tendency is likely to be accentuated as the hardness of the wood and its cut resistance increase. The relationship between the cutting time and the trunk section is probably linked to the magnitude of chain wear. A new chain was used with each fluid to carry out the abovementioned 300 cuts, which resulted in a total cut surface of 9.4 m² for both fluids, while the total cutting times were 2124 s and 2393 s, respectively, for S_B and F₂. The chain wear was evaluated according to the differences between the mass of the new chain and the mass measured after the 300 cuts, after it was accurately cleaned.

Table 16. Evaluation of the wear of the chains after the test with S_B and F₂ in T₃.

Lubricants	State of the Saw Chain	Masses			Specific Chain Wear
		Chains	Residues	Losses	Per surf. Unit	Per Cut	Diff.
		g	mg	mg	mg m−2	mg cut−1	%
SB (viscosity @ 40 °C: 42 mm−2·s−1)	New	172.25	-	-	-	-	-
	End of test: dirty chain	173.74	1490	-	-	-	-
	End of test: cleaned chain	172.06	-	190	20.21	0.63	-
F2 (viscosity @ 40 °C: 47 mm−2·s−1)	New	172.36	-	-	-	-	-
	End of test: dirty chain	174.29	1930	-	-	-	-
	End of test: cleaned chain	172.00	-	360	38.30	1.20	89.47

shows the measurements made on the two chains in the tests with S_B and F₂.

It can be noticed that the mass reduction observed with F₂ (360 mg) was nearly double (1.89 times) compared to S_B (190 mg). At the same time, the mass of the residues detached from the chain by cleaning operations was higher in the chain lubricated with F₂. Here, the residues had a vitrified aspect, and the cleaning resulted in being more difficult. Various factors could have contributed to determine the differences in the levels of chain wear observed with the two fluids. With the same total surface area cut with the two fluids, it must be considered that the trunk with the largest section (535 cm²) was cut with F₂ in Rep. 5 (Table 14), where the cutting times expanded according to the exponential function of Figure 11 and resulted in being much greater than in the other five repetitions. The mechanical and thermal stress conditions that occurred on the chain in Rep. 5 probably were used to make the entire F₂ test, on average, more demanding and to determine the observed higher chain wear with respect to S_B test. As regards the thermal stress with the two fluids, albeit the chain temperatures reported in Table 14 are only indicative and the higher value was observed in Rep. 2 instead of Rep. 5, they seem to confirm that, with F₂, the average thermal level of the chain was higher than with S_B, thus indicating the probable occurrence of a higher friction. Considering the magnitude of the differences between the two wear levels, alongside the considerations reported in the previous point, the greater wear observed with F₂ could partly be derived from an insufficient lubrication, which was highlighted by the severe test conditions, and contributed to the increase in both the friction (between chain parts and guide bar) and the chain temperature [21]. Since new saw chains were used in T₃ and were never sharpened, the observed mass reductions should interest mostly their lubricated parts, like the rivets, the drive links, and the tie straps, while the wear of the cutters caused by the friction with the wood should be limited and similar with both fluids. This point should be verified. The two fluids were used on the chainsaw without modifying the preset flowrate of the chain lubricant. Despite the positive impressions of the operators, who always affirmed that they did not notice any differences in the performance of the chainsaw after replacing S_B with F₂, and considering the higher viscosity of F₂ (Table 7), it is possible that some changes in the flow rate of the lubricant (in the chainsaw, when possible) and/or in the formulation of the fluid led to a reduction in wear, bringing the performance of F₂ closer to that of S_B.

Air Samplings

As in T₂, the residues on the filters of the two air samplers were analyzed to quantify the fractions of inhalable and respirable sawdust and the concentration of metallic elements. Figure 12a shows that the limit of 2 mg·m⁻³ established by the Directive (EU) 2019/130 [24] was always exceeded, except for with the respirable fraction with F₂. Despite the fact that these data relate to a particular context (intensive cutting at fixed point in a restricted area), they seem compatible with the mean concentrations observed during daily chainsaw running in hardwood coppice (Quercus petraea and Fagus sylvatica) [26].

The data relating to S_B and F₂ are compared following the same sequence of repetitions adopted for each fluid: Rep. 1 vs. Rep. 4; Rep. 2 vs. Rep. 5; and Rep. 3 vs. Rep. 6. According to this sequence, the concentrations of both inhalable and respirable particles resulted in being higher with S_B than with F₂ (except in Rep. 1 vs. Rep. 4 for the respirable fraction). The trend is confirmed by considering the average concentration: in both fractions, the mean values resulted in being clearly higher with S_B. This reflects the results of the air samplings made in T₂, albeit there were much higher particle concentrations observed. The reasons why the particle concentration is higher with S_B than with F₂ are not clear. Due to their characteristics, the fluids could interact differently with the sawdust particles, influencing their dispersion in a different way. This will be the subject of further investigation. The concentrations of metallic elements, as reported in Figure 12b,c, are of the same order of magnitude observed in T₁ (Figure 5) during the felling of the Turkey oak wood. Relevant peaks are visible only for Al, whose highest value is observed for F₂ in the inhalable fraction. In the respirable fraction, the Al concentration is higher for S_B. Small increases in K, Ca, Fe, and Ni were observed in both fractions. In general, following the trend of the gravimetric analysis, the concentrations observed in T₃ resulted in being significantly lower than those observed in T₂. Figure 13a provides some information on the dimensional composition of the respirable fraction according to the classification provided by the five-stage impactor. In this case, only one sampling was carried out for each fluid, which lasted the time of the three repetitions.

The total particle concentrations are of the same order of magnitude shown in Figure 12a. Here, the values with F₂ are higher than those with S_B in all classes, conversely to what was observed with the personal air samplers. As described in Section 2.2.4, the five-stage impactor was positioned about 2 m away from the cutting place, while the personal sampler was worn by the operator and therefore was close to the cutting point. Therefore, possible different interactions of S_B and F₂ with the sawdust particles could have affected the dynamics of their dispersion as a function of the distance from the cutting point, causing the different trends of Figure 12a and Figure 13a. The presence of metallic elements in the five classes is shown in Figure 13b for F₂ and in Figure 13c for S_B. Most of them resulted in being below the detectable threshold and were considered to be 0. It can be noticed that, with F₂, the detected elements were Mg, Al, K, and Zn. The concentrations of Al and K are about 20% of the values in Figure 12b,c. With S_B, the highest concentrations were those of Al and Mg, but the presence of Ca was also observed. All said elements were in the lower dimensional class (<0.25 μm), except for K, which was also in the class 0.5–1 μm.

3.2.5. Felling of a Poplar Grove at CREA

This test was carried out in another study concerning the dynamic and energetic requirements of the felling and chipping of a poplar grove. Such an activity represented a chance to gain additional information on the behavior of the ROPO-based fluid, with particular attention paid to the wear of the chain. Therefore, the same ICE chainsaw and the same chain lubricants, S_B and F₂, used in T₂ and T₃ were used in T₄. The poplar grove was divided into two plots, and each plot was felled using one of the two lubricants and a new saw chain. The results of T₄ are reported in

Table 17. Evaluation of the wear of the chains after the felling of the poplar grove (T₄).

Lubricants	State of the Saw Chain	Masses				Cutting Data		Specific Chain Wear
		Chains	Residues	Losses	Trees	Section Mean Area 1	Total Cut Area	per Surf. Unit	per Cut	Diff.
		g	mg	mg	N	cm2	m2	mg m−2	mg cut−1	%
SB	New	171.98				153.72
	End of test: dirty chain	172.59	0.86	250	294		4.5	0.85	55.32
	End of test: cleaned chain	171.73
F2	New	172.04
	End of test: dirty chain	172.88	1.13	290	252		3.9	1.15	74.86	35.33
	End of test: cleaned chain	171.75

¹ The mean section was calculated based on the mean diameter of 13.99 cm (average of total cut diameters).

.

Besides the wear of the two chains consequent to the felling of the respective poplar tree plot, Table 17 also shows some useful data for interpreting the wear itself, like the number of cuts carried out and the mean cut section area. The latter was calculated from the mean diameter resulting from the cut diameters of a sample of 60 trunks taken randomly among the total trunks felled in the two plots. The diameters were measured during the chipping by means of a potentiometric transducer installed on the feed rollers of the chipper (Figure 14) to continuously measure the variation in the diameter of the trunk as it proceeds into the machine towards the chipping organs.

The data of the diameters of the 60 trunks were recorded and processed, selecting, for each trunk, the maximum value, which refers to the cut section. This elaboration provided a mean diameter of 13.99 cm, with a standard deviation of 25.9 cm and a coefficient of variation of 18.5%, which indicates the high variability of the parameter. By hypothesizing that the variability of the diameters (and areas of the sections) of the sample are uniformly distributed between the two plots, multiplying the mean area of cut section by the number of cuts of each plot provides the total cut area. This and the cut number were used to calculate, respectively, the specific chain wear per unit of surface (mg m⁻²) and the specific chain wear per cut (mg cut⁻¹). Despite the fewer cuts made, the chain used with F₂ suffered a higher mass reduction. The data of specific wear emphasize the difference between the two observed wear levels. The difference can be quantified as 35.33%-higher wear for F₂ in relation to S_B. The differences in the wear levels observed with F₂ and S_B are lower in T₄ than in T₃ (Table 16) due to the different hardness values of the wood species used in the cutting tests (respectively, Poplar and Turkey oak). However, both test results indicate that the saw chains lubricated by F₂ underwent higher wear than those lubricated with S_B. The generally higher levels of temperature observed with F₂ in all tests, probably caused by higher frictions, seem to confirm this trend.

4. Conclusions

4.1. Lubricants and Chainsaws Performance

The results shown in this paper provide a positive indication about the possibility of using refined olive pomace oil as a base stock for lubricant for chainsaws, though some aspects concerning its performance should be deepened and improved. The comparison between the ROPO-based formulation (containing 2% thickener and 3 g·kg⁻¹ of antioxidant) and the market-available Stihl Bioplus was made without any modification to the flow rate on the lubrication pump and evidenced, with the former, higher temperatures, probably caused by higher friction both among the various elements of the saw chain and between these and the guide bar, which resulted in higher chain wear than with the latter [27]. This wear does not seem to affect the efficiency of wood cutting, as testified by the positive opinions of the skilled operators involved in the tests who never noticed any difference between biobased and conventional fluid. However, it would be advisable to check whether it could in any way jeopardize the life of the chain (linked to the wear of the teeth following sharpening) and, consequently, the operator’s safety conditions. Based on the experiences described above, a further series of comparative tests would help to clarify doubts about the suitability of ROPO as a lubricant for chains, suggesting a way to improve its performance by bringing it closer to that of the conventional fluid. These tests should be focused on the “systematic” acquisition of the following data:

-

Measurement of the temperature of the fluids within the reservoir during the cutting.

-

Measurement of the temperature of the saw chain/guide bar system and of the wood by means of the thermal imaging camera during the cutting with both fluids.

-

Saw chain lubricant consumption during cutting with both fluids and evaluation of any differences. e.g., F₂, used from T₁ to T₄, is more viscous than S_B, and its consumption should therefore be lower and cause more friction, heating, and wear of the chain.

-

Based on the results of the previous point, an ROPO-based formulation with a different viscosity could be tested to reduce the difference in chain wear relating to the commercial lube. Another possibility to investigate is to change the flow rate setting of the pump in the chainsaw’s lubrication system in order to optimize the lube distribution on the chain.

-

Evaluation of the long-term stability of the biofluid, both in the storage phase and after sitting for a long time in the unused chainsaw.

To compare the results provided by the lubricants, the data should come from tests at a fixed point, which would allow us to maximize the repeatability of conditions like trunk dimensions, ambient temperature, and relative humidity; and to standardize the dust sampling and the thermal monitoring of the chain, lube, and wood.

4.2. Lubricants and Sawdust Emissions

The results of the air samplings and analyses of the residues revealed that the concentrations in the air of the inhalable and respirable fraction of wood dust particles both often exceed the limit of 2 mg·m⁻³ established by the Directive (EU) 2019/130 and, thus, may increase the risk of cancer from workers exposure to this material. It was also found that dust particles contain variable amounts of metallic elements (like Al, K, Ca, Cu, and Zn), and these resulted in also being present in the respirable fraction. These concentrations were found to be lower with the ROPO-based fluid than with the conventional fluid, and the reason for this needs to be clarified: it could depend on some different interaction of the wood dust particles with the dispersed fluids droplets. The sawdust also acts as a vector of chain lubricant and other substances (like metallic elements) dispersed by the chainsaw during cutting, the presence of which in the respirable fraction makes them potentially dangerous for the health of the operators, increasing the risk deriving from exposure to mere sawdust. In a study on the dispersion of chainsaw oil based on the use of radiotracers, it was determined that 21.9% falls onto the ground, 8.3% is soaked into the cut trunk’s surface and into the immediately surrounding area, about 5% remains in the air as aerosol, and, finally, 64.4% of the total lubricant consumption is retained by sawdust [9]. Therefore, wood dust particles, as the lighter part of the sawdust, will also probably contain oil that is capable of deeply penetrating the respiratory system. Its amount could be assessed by combining the just-cited data and some results of our study. For example, relating to the results of T₁, the assessed amount of oil dispersed per surface unit is 28.8 L ha⁻¹ of felled forest. Of these, according to the aforementioned percent distribution, 1.6 L ha⁻¹ would be in aerosol form, and 20.2 L ha⁻¹ would be incorporated in the mass of sawdust. Considering the number of cuts per hectare, the mean area of cut section, the thickness of the cut (Table 9), and a mean specific gravity of 900 g L⁻¹ for the wood of Turkey oak, the mass of sawdust can be assessed at around 3145 kg ha⁻¹, with a resulting mean concentration of oil in sawdust of 0.32%. Hypothesizing that this remains constant in the dust wood particles, in the case of a concentration of respirable fraction of around 3 mg m⁻³ (Figure 12a, S_B), the oil it contained would be 9.65 mg m⁻³. This value probably underestimates the actual respirable oil amount because it does not comprehend the aerosol fraction, and, most of all, the operators constantly work near the cut point, where the wood dust concentration is higher. As regards the effects of these values on workers’ health, the Directive (EU) 2019/130 refers to hardwood dust as a carcinogenic risk factor. Other substances present in wood dust, like metallic elements (arsenic, lead, cadmium, and nickel), could increase the risk from exposure to simple wood dust [22,28].

From this point of view, the risk is probably affected by the presence of the exhausts of two-stroke-engine chainsaws, which are characterized by the presence of mutagenic substances due to the combustion of the gasoline–lubricant mix [6]. According to the same approach and data used to assess the concentration of chain lubricant in sawdust, the consumption of mix per surface unit is 52.5 L ha⁻¹, which contains 1.6 L ha⁻¹ oil (at 3% in the mix). However, the results of T₂ did not evidence the influence of exhausts in the concentration of wood dust and metallic elements. It would be interesting to investigate the presence, in the residues from air sampling, of organic substances relevant to the health of the operators, their origins (from chain lubricant, additives, combustion exhausts, etc.), and any synergistic effects with those of wood dust. To this purpose, the use of WEM chainsaws and ICE chainsaws of comparable performance in specifically designed tests would be helpful and would allow us to improve the knowledge of the quality of the emissions with conventional and biobased lubricants. In any case, the use of highly degradable lubricants for saw chains would reduce the contribution of this factor to the global risk of exposure to all emissions from the use of chainsaws. Lubricating the saw chains with biobased lubricants like ROPO would ensure a high level of biodegradability, which should be safeguarded through a correct choice of additives, as in the case of the antioxidants adopted in our tests. Most of these considerations can be extended to the aspect of the impact of the activities based on the use of chainsaws on the environment—in particular, forests, which receive a high load of pollutants per surface unit.

The above-described process should lead to the realization of a chain lubricant that is technically valid and environmentally sustainable. Its introduction into the operative reality will depend on future market trend. The present study began in 2021, before the start of the Ukrainian war (22 February 2022), which is still ongoing and has caused, among other things, strong disruptions in the market of all vegetable oils, whose prices have increased dramatically. In relation to our study, the ROPO, previously considered a by-product, acquired new value as edible oil, and its price increased from 1.90 EUR kg⁻¹ in January 2022 to 3.10 EUR kg⁻¹ in October 2022 [29] and to 3.40 EUR kg⁻¹ in October 2023 [30], making it non-competitive with conventional chain lubricants. The desirable end of the war could help restore market conditions and make them more favorable to the use of products such as ROPO in non-food applications.

The results obtained in this research on ROPO can be transferred to other vegetable oils with a similar fatty acid composition, such as high-oleic sunflower oil and rapeseed oil, when the price and/or availability of ROPO oil should be critical. The only remarkable difference for the discussed applications stands in the diglyceride content of these refined oils, as the acidity of the corresponding crude oils is lower.