FIREFIGHTING COMBAT WITH FIRE RETARDANTS AT DIFFERENT CONCENTRATIONS

Forest fires are responsible for the destruction of millions of hectares of forest worldwide, and they lead to diverse economic, social, and landscape damage. Thus, the development of techniques to combat them has become increasingly necessary. In this context, this study aims to evaluate the efficiency of different fire retardants at different concentrations in fighting forest fires, considering the burning times and intensities of forest fuel. The study was conducted inside Eucalyptus spp. stands using three fire retardants (Silv-Ex, F-500, and HoldFire) at three concentrations (1%, 1.5%, and 2%), in addition to a water-only control. A completely randomized design was used, and the statistical analysis was completed based on experimental arrangements (factorial 3x3). Variables evaluated during the burning process were as follows: burning times (the time required for the flames to consume all forest fuel within the sample, with and without the retardant) and intensity of burning. Results regarding the time and intensity of burning in relation to the concentrations indicated a decreasing trend as the latter were increased, classifying the highest dose (2%) as the most efficient. For the retardants, all were observed to be efficient, with Silv-Ex being the most appropriate as it significantly reduced the burning intensity and increased the burning time of the forest fuel.


INTRODUCTION
Forest fire can be understood as any uncontrolled fire that occurs in vegetation and is either anthropic or natural in origin. In general, forest fires are responsible for the destruction of millions of hectares of forests worldwide and generate widespread economic, social, landscape, and soil damage, which may occur in reforestation areas, conservation units, preservation areas, farms, roadsides, and surrounding urban areas (WESTERLING; BRYANT, 2008;SOUTO et al., 2009;TORRES et al., 2011;BOSCARDIN et al., 2014;McKENZIE et al., 2014;KNELMAN et al., 2017;RODRÍGUEZ et al., 2017).
In Brazil, fires are classified as the biggest catastrophe that a forest stand can suffer, as they cause economic damage, contribute to climate change, and endanger human and animal lives. (DINIS; CARVALHO, 2011;SOUZA et al., 2012). The National Institute for Space Research (INPE) recorded an annual average of 38,609 outbreaks of fires in the Brazilian territory from 2012 to 2018, and statistical evidence shows that the number of fires tend to increase as a result of unusual or cyclical global climate changes related to El Niño and La 1108 FLORESTA, Curitiba, PR, v. 50, n. 1, p. 1107-1112, jan/mar 2020Fiedler, N. C. et.al Electronic ISSN 1982 Niña and from the indiscriminate and irrational use of fire by humans (BATISTA, 2009;TETTO et al., 2015;TORRES et al., 2016;INPE, 2017).
Although there are numerous preventive and firefighting techniques, these do not solve or prevent all occurrences (FIEDLER et al., 2015). Thus, the improvement of such techniques is always of high importance to eradicate or at least minimize the damage caused by the conflagrations.
The evolution of forest firefighting methods highlights an increase in the use of chemical products incorporated in water to enhance their efficiency, called flame or fire retardants (BATISTA, 2009). The potentiating characteristics of fire retardants are extremely important because of the difficulty of obtaining water in the regions of fire occurrence, and the favored ones provide this function at a relatively low cost (CANZIAN et al., 2018).
Brazil is not yet endowed with legislation that regulates the use and quality of these products, or even advises users in the correct methods and quantities to be used (PIERONI et al., 2017). One of the biggest limitations on the use of fire retardants in Brazil is high prices, as the vast majority of them are imported.
From the above information, it is evidently necessary to study different retardants and the possibility of maintaining the quality of their performance at lower concentrations, to maximize the effects of water with such products and to reduce costs. Thus, the objective was to evaluate the effectiveness of combatting forest fires using the fire retardants in different concentrations. The hypotheses that guide the present study are as follows: a) the natural base retardant used in the present research will present statistically inferior results to the chemical ones; and b) higher product concentrations will ensure more promising results.

Study area
The study was conducted in an experimental area of the Federal University of Espirito Santo, located in the municipality of Jerônimo Monteiro, southern Espírito Santo State, Brazil, between the coordinates 20°47'45"S and 41°24'21"W. The climate is classified, according to Köppen, by the Aw type, whose average annual temperature is 23 ºC and average rainfall of 1200 mm year -1 (INCAPER, 2013).

Fuel material used
In all plots, 2.5 kg m -2 of dry Eucalyptus spp. included in floor-mounted windrows ( Figure 1) was used as standardized fuel material to be treated with the retardants. The moisture and dry matter quantification of the fuel material was determined by oven drying (65 ºC) it until it reached a constant weight. Subsequently, a sample of 3 kg of forest fuel was allocated to a template of 1 m 2 .
The moisture content of the fuel material was determined by a three-repetition sample of 1 m 2 of the fuel material, which corresponded on average to 2.95 kg m -2 wet litter.

Methods used
Each plot was subjected to a controlled application of 0.5 L of solution or water (control) per square meter (m²), as proposed by Fiedler et al. (2015), with retardant treatments differentiated by type and concentration (1%, 1.5%, and 2%). The concentration recommended by the manufacturers is 1%; however, it was decided that higher FLORESTA, Curitiba, PR, v. 50, n. 1, p. 1107-1112 concentrations would be applied to determine whether the efficiency of the products would increase proportionally with the consequent reduction in the amount of water. This is especially important considering the difficulty in obtaining water at the time of firefighting. The retardants used in this study were as follows: • Silv-Ex (SV): A liquid used for Class A fires, with recommended concentrations ranging from 0.1% to 1% in fresh, brackish, or saltwater; • F-500 (F5): A liquid with recommended concentrations that can range from 0.5% to 3% according to fire class (Classes A, B, and D); and • HoldFire (HF): A biodegradable additive that enhances water use in Class A fires.
With the aid of an adjustable sprayer, the solution containing the fire retardant was mixed with water to achieve the desired concentration in the intended spaces (1 m x 1 m), and then was homogeneously applied before combustion began.

Sampling procedure
The experimental area consisted of 30 plots, with dimensions of 1 m x 3 m. Within a plot, the retardant application area was 1 m x 1 m, and the ignition area of the fire that defined the direction of fire was the remaining 2 m x 1 m ( Figure 2). In total, there were nine treatments with three repetitions each. The other plots were the control plots treated with water only. After the application of the retardants, each fire was ignited in the forest fuel so that the flames followed the direction of propagation until they reached the end that was subjected to the treatments. The first part of the windrows only served to form the line of fire, while the distance, used for calculation purposes, was accounted for only in the second part.

Fire characterization
Relative humidity, wind speed, and plot burning time with and without the product were evaluated. The characterization of fire behavior was performed using the burning intensity equation (1), proposed by Byram (1959).

Statistical analysis
For statistical analysis of time and intensity values, the experimental arrangement was factorial, considering the types of retardants and their concentrations, the control with water (treatments), and three plots as repetitions. Initially, the analysis of variance (ANOVA) was applied, and subsequently, for the comparison of means, the Tukey test was applied at 5% probability of error with the aid of the SISVAR 5.6 program.
Using Excel 2013 software, it was verified, based on linear regression, that a functional relationship existed between the independent variables (product concentrations) and the dependent variables (burning times and intensities).

RESULTS
The mean values and their respective standard deviations of mass, humidity, height, and density of the forest fuel are presented in Table 1.  Table 2 presents the results of the analysis of variance (ANOVA) for the burning times and intensities. The ANOVA results for burning times demonstrated the existence of significant interaction between the main factors (retardants and doses) by Tukey test at 5% probability of error, which led to the need to proceed with the unfolding of such factors (Table 3). The analysis of variance for the burning intensity showed no significant interaction between the main factors (retardants and doses) by Tukey test at 5% probability of error (Table 4).  Figure 3, it is seen that as the retardant concentration increases, the burning intensities decrease, and the burning times increase.

DISCUSSION
The amount of dry forest fuel used in the present study was 20.9 Mg ha -1 . This value meets the reality of Brazilian planted forests, because according to Souza et al. (2003), the amount of surface forest fuel in a 7-yearold Eucalyptus dunii Maiden forest approached 22 Mg ha -1 . Already, the density of the material, when compared to the results found by White et al. (2014) in eucalyptus forests on the northern coast of Bahia State, Brazil, was considered low, as they found it to be 19.95 kg.m -3 , approximately 17% more than that found in this research (16.66 kg.m -3 ).
The results on the intensity of burning were very interesting, since the expected fire retardants and higher concentrations led to a decrease of these values. Regarding the concentrations, the concentration of 2% emerged prominent, as it reduced the burning intensity by approximately 53% in comparison to the control with only water. These results corroborate those found by Canzian et al. (2016), who obseved an inversely proportional relationship between retardant doses and burning intensity.
The retardant that obtained the best results on burning intensity was the SV, showing a reduction close to 54% compared to the control with water alone. It is also noteworthy that HF is an organic-based retardant that is little-known in the market, and it was able to reduce the intensity of burning by 51% compared to the control. Similar results were found by Maraboti et al. (2016) when studying the intensity of burning in forest fuels subjected to different fire retardants. The authors found, SV to be the best treatment, coincidentally, the F5 to be the worst.
Regarding the burning times, the trend remained as expected, with the highest doses producing the best results. However, it is noteworthy that SV was more effective at all doses, and only at 1% dosing did it match the HF results. The SV showed its best results in the highest dosage (2%), presenting about 102% increase in the burning time compared to the control without retardant. The HF presented similar results in the doses of 2% and 1.5%, presenting increase in the burning time in relation to the control, in the class of 125% and 124%, respectively. The F5 was more efficient in the dosage of 2%, which improved the burning time by 101% compared to its control.

CONCLUSIONS
• Among the three retardants studied and the water-only control, the treatment that presented the best results, at all concentrations, both for time and intensity of burning, was Silv-Ex; therefore, it was considered the most efficient under the conditions analyzed; • Concentrations tended to decrease burning intensity and increase burning time of the forest fuel as they were increased, rendering the highest dose (2%)  • As organic-based retardant is a newly developed product, it was quite effective in combating forest fires given the results presented.