Glycerin effluent from the biodiesel industry as potassium source to fertilize soybean crop

1 Faculdade de Ciências Agrárias, Universidade Federal dos Vales do Jequitinhonha e Mucuri (UFVJM), 39100-000 Diamantina, Minas Gerais, Brazil. 2 Programa de Pós Graduação em Produção Vegetal (UFVJM), Faculdade de Ciências Agrárias, 39100-000 Diamantina, Minas Gerais, Brazil. 3 Universidade Federal dos Vales do Jequitinhonha e Mucuri (UFVJM), 39100-000 Diamantina, Minas Gerais, Brazil and Departamento de Química ICEx, Campus Pampulha, Universidade Federal de Minas Gerais (UFMG), 31270-901 Belo Horizonte, Minas Gerais, Brazil. 4 Departamento de Engenharia Agrícola, Universidade Federal de Lavras (UFLA), 37200-16 000 Lavras, Minas Gerais, Brazil.


INTRODUCTION
The residual crude glycerin from the biodiesel industry corresponds to about 10 to 15% of the total biodiesel mass production. In the lack of any specific legislation indicating how to dispose such residue, much of this by-product is more commonly accumulated in areas of the industrial plants. The expected expansion of the worldwide production of biodiesel will in consequence tend to increase the stock of glycerin (Ooi et al., 2004).
The residual crude glycerin from the biodiesel industry usually contains about 50 mass% of pure glycerin (Zhou et al., 2008;Carvalho et al., 2012) and is rather dark, relatively to the pure glycerin; it needs to be purified for further use in the fine chemistry industry (Yong et al., 2001). However this process is expensive and the effluent is usually discarded as waste glycerin by small industries producing biodiesel (Dasari et al., 2005).
Following the increasing production of biodiesel worldwide, the glycerin waste is thought to be a potential environmental problem although it may also represent an opportunity to get environmental and economical profits depending only on new technological developments favoring its rational use. Many challengeable alternatives may be put under considerations, in attempts to confer better technological uses, reduce the environmental impacts and consolidate biodiesel as an environmentally and economically competitive biofuel (Ito et al., 2005). Some technological uses of crude industrial glycerin as feedstock have been reported (Zhou et al., 2008), although its potential use as soil fertilizer for agricultural lands is still unknown.
Soybean oil is the most widely used for the production of biodiesel (Freedman et al., 1986;Noureddini and Zhu., 1997). Thus the lower the cost of the soybean production the more interesting would be its industrial use. A conceivable alternative for the agricultural use of crude glycerin from the transesterification of triacylglycerides of bio-oils to produce biodiesel, using potassium hydroxide as catalyst, would be to neutralize the effluent with phosphoric or sulfuric acid. The potassium phosphate or sulfate formed during this step would improve the use of the effluent as fertilizer (Zhou et al., 2008).
Potassium is the second most promptly absorbed macronutrient after nitrogen. The plant nutritional requirements may be variable, depending on innumerous conditions of soil and the plant itself. On average, it assumes an uptake of 81 kg N and 54 kg K to produce 1,000 kg of soybean grains (Borket and Yamada, 2000). Some studies have shown that when levels of available potassium in soil are above 60 mg dm -3 , the yield responses of the soybean plant to potassium fertilization are usually not significant (Scherer, 1998a, b). Maximum yields of soybeans were reportedly attained with application of 60 kg ha -1 (Scherer, 1998b); 80 and 120 kg ha -1 , if the level of available K in soil was between 16 and 40 mg dm -3 (low content) and between 41 and 70 mg dm -3 (mean content), respectively (Novais, 1999) and 85 and 90 kg ha -1 K 2 O in case of no-tillage land management (Foloni and Rosolem, 2008).
In tropical and subtropical soils, organic matter has a close relationship with other physical, chemical and biological soil properties, chemical among the cation exchange capacity (CEC), is of fundamental importance for maintaining the productive capacity of the soil longer term (Ciotta et al., 2003;Araújo et al., 2007;Moreti et al., 2007;Carvalho et al., 2011). Thus, industrial glycerin, as an organic residue, would tend to increase the usually low CEC of tropical soils, to reduce loss of cations, particularly K + , and improve their fertility. In this work, it was used sulfuric acid to neutralize the material of the glycerin effluent. The resulting raw product containing potassium sulfate was assayed as fertilizer for soils supporting the growth of soybean plants. The study thus aimed at evaluating the use of the glycerin effluent (the material will be hereinafter referred to as "Kglycerin"), a waste product of the biodiesel industry, from the transesterification of triacylglycerides in a vegetable oil catalyzed with KOH, as a potassium source on the yield and nutrition of soybean plants and on alterations of the chemical and microbiological properties of the soil. The soil samples were air-dried, sieved (2.0 mm) and characterized according to Embrapa (1997) (Table 1). The pH was potentiometrically measured (soil:water 1:2.5, v/v); phosphorous and potassium were extracted with the Mehlich-1 solution and determined by colorimetry (for phosphorous) and flame photometry (potassium); calcium, magnesium and aluminum were extracted with a solution of 1 mol L -1 KCl and determined by flame atomic absorption spectrophotometry (calcium and magnesium) and titrated with 0.025 mol L -1 NaOH, for aluminum; the acidity (H + Al) was determined by extracting with 0.5 mol L -1 calcium acetate buffered at pH 7.0 and quantified by titration with 0.025 mol L -1 NaOH. The organic carbon (OC) was determined by Walkley-Black method (Walkley and Black, 1934). The sulfate (SO4 2-) content was determined according Hoeft et al. (1973). Values of cation exchange capacity (CEC), Al saturation (m) and base saturation (V) were then calculated. The pipette method was used to determine the soil texture (Embrapa, 1997).

Planting and agronomic practices
The preparation of the experimental area was made by the *Corresponding author. E-mail: ebsilva@ufvjm.edu.br Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License conventional system with disc plow to 0.4 m deep and two disking with disc harrow. Liming was carried out three months before planting, by distributing the powder dolomitic limestone (90% total neutralizing power) all over the area with; the material was incorporated at 0.30 m depth at a dose corresponding to 2.4 x 103 kg ha -1 , for both soils. The soybean population of the cultivar MSOY-8001, during the spring of 2008; the seeds were manually spread over to get a stand of 444,444 plants ha -1 , with spacing of 0.45 m between rows and 20 seeds per meter along rows. The phosphorus fertilization was done at the planting time at a dose 120 kg ha -1 P2O5 as superphosphate (18 mass% P2O5 and 10 mass% S) in all treatments and cultivation places. The seeds were inoculated with concentrated liquid inoculant (125 mL of liquid inoculant to 50 kg of seeds) o Bradyrhizobium japonicum. All cultural practices were followed as recommended by Embrapa (2008).

Treatments and experimental design
The crude glycerin was obtained from the conventional chemical industrial process of transesterification of triacylglycerides, in this case, of the cooking soybean oil after being used in frying foods. The transesterification was catalyzed with potassium hydroxide and the glycerin effluent was finally neutralized with sulfuric acid. The resulting liquid material will hereinafter be referred to as "Kglycerin". The chemical composition for this "K-glycerin" (pH = 8.3 and density = 1.01 kg dm -3 ), as could be determined according to analytical procedure described by Melo and Silva (2008), in kg m -3 , is: P2O5 = 0.5; K2O = 24.9; S = 1.1; B = 1 x 10 -3 ; Cu = 4 x 10 -3 ; Fe = 54 x 10 -3 ; Mn = 2 x 10 -3 ; Zn = 96 x 10 -3 and organic carbonic = 3 x 10 -3 .
The field experimental design was a randomized block design with four replications. Each experiment plot consisted of four 3.0 mlong rows, spaced of 0.45 m, comprising a total area of 5.4 m 2 . The useful area of each plot was considered by taking two central rows, excluding 1.0 m from each row-end.
The treatments consisted of four K2O doses on the TH (0; 40; 80 and 160 kg ha -1 K2O) and TQ (0; 60; 120 and 240 kg ha -1 K2O) sites. These K2O doses correspond to 0; 1.6; 3.2 and 6.4 m 3 ha -1 "K-glycerin" on TH, and 0; 2.4; 4.8 "K-glycerin" and 9.6 m 3 ha -1 on TQ. Two additional treatments were made to supply inorganic sources of potassium (KCl, corresponding to 58 mass% K2O, and K2SO4, 48 mass% K2O) at a dose of 80 and 120 kg ha -1 K2O on TH and TQ, respectively. These doses of potassium were based on recommendations for the soybean crop (Novais, 1999) and on the availability of potassium in the soil, evaluated according to standard procedures recommended for the soil chemical analysis (Table 1). The liquid "K-glycerin" was manually distributed all along the sowing rows at a depth of 7 cm.

Inputs and measurements
To evaluate the soybean crop yield, the grain moisture was corrected to 13 mass%. After harvesting the soybean grains, the soil samples were taken at a depth between 0 to 0.20 m from the soil surface, for each plot, in order to determine all other chemical properties. About 20 soil samples were collected from each plot in order to prepare one mixed sample. There were determined the pH in water, P and K (Mehlich-1 extractor), Ca and Mg (KCl 1 mol L -1 extractor), H + Al (calcium acetate 0.5 mol L -1 extractor) and the organic carbon (OC) according to the Walkley-Black method (Embrapa, 1997); the sulfate (SO4 2-) content was determined according to the procedure described by Hoeft et al. (1973). Values of cation exchange capacity (CEC = Σ(K, Ca, Mg, H + Al) in mmolc kg -1 ) and saturation in bases (V% = [Σ(Ca, Mg, K)/(CEC)] × 100) were then calculated.
The nutritional status of the soybean leaf samples was determined at the stage R2 (full flowering) per useful plot, in order to determine the nutrient composition. The amount of nitrogen was determined by the semi-micro Kjeldahl method (Cunniff, 1995). The P, K, Ca, Mg, S, Cu, Fe, Mn and Zn contents were obtained by nitric perchloric acid digestion (Miller, 1998). The contents of Ca, Mg, Cu, Fe, Mn, and Zn were determined by atomic absorption spectrometry; K was determined by flame photometry (Isaac and Kerber, 1971). S was determined by the barium sulfate turbidimetry (Beaton et al., 1968). The B content was determined by colorimetry (azomethine method) after dry digestion (incineration) (Wolf, 1974).
Ten soil sub-samples were randomly collected at a depth between 0 and 0.10 m deep. The first emergence soybean plants were collected, packaged in plastic bags and transported in thermally isolated boxes. The materials were sieved (2 mm) in order to remove organic residues and roots. Then sub-samples were stored at 4ºC until microbial analyses were made. Flurescein diacetate hydrolysis (FDA) by soil microorganisms was evaluated according to Frighetto and Valarini (2000). Soil basal respiration (Rbasal) was estimated based on CO2 released from four subsamples (20 g each) taken from each mixed soil sample (water content, 60 mass%). The samples were sealed in a 1.0 L flask with 10 mL of 0.3 mol L -1 KOH and titrated, after three days, with 0.1 mol L -1 HCl, according to the methodology described by Alef and Nannipieri (1995). Soil microbial biomass C (Cmic) was determined by fumigation and incubation, as described by Jenkinson and Powlson (1976). The metabolic quotient (qCO2) was determined by the ratio between Rbasal and Cmic (Anderson and Domsch, 1993), expressed in μg CO2 μg Cmic -1 day -1 .

Statistical analysis
The numerical data were subjected to analysis of variance for the following factors: blocks, soil type, K2O doses as "K-glycerin" and inorganic sources of potassium (KCl and K2SO4), as an additional nutrient supply. The microbiological analysis of soil, and the factors above, added two evaluation periods (early emergence and harvest) of the soybean. The separation of the mean values was done using the criterion of the least significant difference (LSD) at a 5% probability level. The fitted equations for the variables evaluated in terms of K2O doses applied as "K-glycerin".   1. Grain yield of soybeans as a function of K2O doses in the form of "K glycerin" (KG) and K inorganic sources (KCl and K2SO4) into two types of soil (*Significant at 1% by F-test).

RESULTS AND DISCUSSION
Grain yield of soybean influenced by the K 2 O doses in the form of "K glycerin" to increase quadratically when cultivated in soil TH and linearly when cultivated in soil TQ (Figure 1). The inorganic sources (KCl and K 2 SO 4 ) applied in TH did not differ and when applied in TQ were superior and differed doses of K 2 O in the form of "K glycerin" (Figure 1). The K 2 O doses in "K glycerin" in applied TH caused reductions in soybean yield due to nutritional imbalance caused by competition cationic and anionic nutrients (Parker and Norvell, 1999). The lack of increase in grain yield of inorganic sources relative "K glycerin" and the latter by reduction with increasing K 2 O doses in TH due to the initial K content in soil (Table 1), with a value above 40 mg kg -1 of soil (Rosolem et al., 1993;Scherer, 1998a;Borket and Yamada, 2000). Differently than soybean cultivation in TH, response to K fertilization with application of "K glycerin" and inorganic sources occurred when exchangeable soil K was below 40 mg kg -1 in TQ soil. Responses to K fertilization in soybean were found in Typical Hapludox with K content of 27.3 mg kg -1 of soil (Mascarenhas et al., 2000). The maximum yield (1,157 kg ha -1 ) of soybean attained with the maximum dose applied K 2 O (240 kg ha -1 ) in the form of "K glycerin" on the ground TQ (Figure 1). In soil TH, maximum yield of 1,622 kg ha -1 attained with 87 kg ha -1 of K 2 O with "K glycerin" (Figure 1). The K 2 O dose were higher than the determined by Scherer (1998b) and the near recommended for soybean (Novais, 1999) and obtained in succession millet-soybean no-tillage (Foloni and Rosolem, 2008) to achieve maximum productivity in TH soil. In TQ soil, the K 2 O dose as "K glycerin" to achieve maximum yield was more elevated than in TH soil.
The yield of soybean crop in Brazil in 2012/2013 was 2,938 kg ha -1 and the Minas Gerais State, Brazil was 3,010 kg ha -1 (Conab, 2013) with fertilizer application in soybean seeding. The conditions of this study, the yields on both cultivated soils were significantly lower than the national and regional average. The results attributed to climatic conditions during the experimental period with a mean temperature of 24.2 and 20.0°C and rainfall of 735 and 710 mm in Curvelo and Diamantina, respectively. The favorable climatic factors are essential for proper yield of soybean with water requirement of 7 to 8 mm per day, totaling in cycle from 450 to 800 mm, depending on the duration and management of the crop cycle and optimal temperature around 30°C (Embrapa, 2008). The temperature interfered on soybean yield in both cultivation places, with greater effect in Diamantina, besides the difficulty of K fertilizer management due to the low CEC in TQ soil (Table 1), which provides low yield with higher K dose applied to the soil. The need for application of dose above 60 kg ha -1 of K 2 O in soils with low CEC and in areas subject to intense rainfall, K should be applied broadcast at doses allowing maintenance of adequate levels in soil and that return the quantities exported by crops (Guareschi et al., 2008).
The direct application of "K glycerin" was evaluated as an alternative source of nutrients for the soybean crop, thus the soil chemical attributes and levels in the leaves Table 2. Effect of K2O doses in the form of "K Glycerin" (KG) and inorganic sources (KCl and K2SO4) on soil chemical attributes applied to two types of soil. of soybean was evaluated (Tables 2 and 3). The K and S-SO 4 2contents in soil affected quadratically (K: ŷ = 75.2 + 1.19x -0.006x 2 , R 2 = 0.95 and S-SO 4 2-: ŷ = 13.4 + 0.35x -0.002x 2 , R 2 = 0.98) in TH and linearly in TQ (K: ŷ = 19.0 + 0.15x, R 2 = 0.98 and S-SO 4 2-: ŷ = 19.9 + 0.05x, R 2 = 0.99) by increasing K 2 O doses of "K glycerin" that differed from K inorganic sources (KCl and K 2 SO 4 ) applied (Table 2). In contrast, the K in soybean leaves only was affected linearly (K: ŷ = 7.2 + 0.0171x, R 2 = 0.92) by increasing K 2 O doses of "K glycerin" in TQ which differed from K inorganic sources applied (Table 3). Foliar S increased with K 2 O doses of "K glycerin" linearly in TH soil (S: ŷ = 0.8 + 0.0046x, R 2 = 0.95) and in TQ soil (S: ŷ = 1.2 + 0.0016x, R 2 = 0.89) and differed from K inorganic sources in both soils (Table 3).
The diagnosis of a nutrient deficiency, it is important to conduct soil testing in advance, to make possible corrections in the fertilization to minimize future losses in yield. Soybean yield (Figure 1) had a relationship with the available K and S content in the soils ( Table 2). The exchangeable soil K in TH by initial analysis (Table 1) is above 40 mg K kg -1 soil whereas the TQ is below (Rosolem et al., 1993;Scherer, 1998a;Borket and Yamada, 2000). In contrast, the availability of the S-SO 4 -2 in the initial analysis (Table 1)

kg
-1 soil (Huda et al., 2004) in both soil cultivated with soybeans. This S and K availability in the soil reflected in the content of such nutrients in the leaves of soybean (Table 3). The contents of K and S in the soil increased with "K glycerin" doses since these nutrients constituents of this waste arising from the production process of biodiesel used, with the K inorganic sources and due to P fertilization of soybean with superphosphate.
The increases in S and K in the soil (Table 2) as reflected in contents of these nutrients in the leaves of soybean (Table 3) provided no increases in grain yield in TH (Figure 1), due to S content in the leaves of soybean being below the 2.5 g kg -1 amount (Martinez et al., 1999) and 2.3 g kg -1 (Urano et al., 2007). The nutrient with the greatest increase in TQ both in soil (Table 2) as in the leaves of soybean (Table 3) was the K thereby could be responsible for increased yield. The K influences various physiological processes such as photosynthesis, transport of photoassimilates and enzymes activation, which directly affected the yield (Pettigrew, 2008). The K content in the leaves below 14.0 g kg -1 (Scherer, 1998a), of 17.0 g kg 1 (Martinez et al., 1999) and 23.1 g kg -1 (Urano et al., 2007) related to low yield of soybeans ( Figure 1) at the maximum dose of "K glycerin" in TQ.
The effect of K in Ca and Mg uptake, which normally interact with this nutrient (Mascarenhas et al., 1988), decreased the levels of Ca and Mg in the leaves of soybean when cultivated in TQ (Table 3). The difference of Ca and Mg in TQ may be due to competition with K, since they use the same absorption sites (Andreotti et al., 2001), and the increase of K intensifies competition with Table 3. Effect of K2O doses in the form of "K Glycerin" (KG) and inorganic sources (KCl and K2SO4) on nutrient concentrations in the soybean leaves applied to two types of soil. Ca and Mg (Oliveira et al., 2001). In TH, the high content of K in the initial analysis (Table 1) probably equated to the Ca and Mg competition occurring balanced as opposed to TQ, with higher competition between the exchangeable bases. Despite the increase in K and S in soil (Table 2) and in leaves of soybean (Table 3), application of the "K glycerin" was not sufficient to modify the CEC and the organic carbonic content in two soils (Table 2). A single application of the "K glycerin" in planting furrow added at a dose of 240 kg ha -1 K 2 O in TQ a quantity of 28.9 kg ha -1 of organic carbonic, being required to maintain the initial stock carbon organic soil a total annual addition of 8,900 kg ha -1 carbon in tillage cropping system (Lovato et al., 2004). The increase in organic matter and consequent increase of soil CEC by addition of organic waste was obtained when applied to the total area (Carvalho et al., 2011) and the use of organic waste (poultry manure) did not affect the organic matter in depth up to 0.20 m (Moreti et al., 2007), while the organic matter effect restricted to the surface layers of the soil, which not reflected in the 0 to 0.20 m soil due to the dilution effect on the soil mass (Ciotta et al., 2003;Araújo et al., 2007).
Microbiological analyzes were performed at two different times in order to obtain information regarding the direct application of "K glycerin" in soil microbes (Table  4). The reduction of microbial activity measured by flurescein diacetate hydrolysis (FDA) with increasing K 2 O doses (FDA: ŷ = 4.36 -0.019x + 0.00012x 2 , R 2 = 0.83) in the form of "K glycerin" in TH (Table 4) and the source K 2 SO 4 may have occurred because the dose was detrimental to soil microbes at the time of soybean emergence (E1). The addition of organic and inorganic fertilizers can cause positive or negative effects on microbial biomass and its activity (Böhme et al., 2005).
It was observed reduction value of the FDA at the time of harvest of soybean (E2) in relation to the time E1 did not differ between the application of "K glycerin" doses and K inorganic sources in TH (Table 4). Already in the soil TQ, microbial activity was not influenced by the levels of "K glycerin" and K inorganic sources at the time E1 and an increase in the time that E2 (FDA: ŷ = 1.98 + 0.002x, R 2 = 0.82) did not differ from K inorganic sources (Table 4). The evaluation times, the behavior was different in relation to TH because TQ on the end of the experimental period the value was greater than the initial fact justified by the non-occurrence of drought stress in the experimental area. This may be the effect of climatic conditions, where there was a moisture stress in the experimental period which may have affected the soil microbes, as adverse environmental conditions affect the soil microbial activity (Nannipieri et al., 2003).
In soybean emergence (E1), with the addition of "K glycerin", microbial basal respiration (Rbasal) increased linearly in TH (Rbasal: ŷ = 1.32 + 0.015x, R 2 = 0.85) and quadratic (Rbasal: ŷ = 1.15 + 0.034x -0.00011x 2 , R 2 = 0.94) in TQ, which only differed between the two K inorganic sources in the last soil (Table 4). The application of organic waste ("K glycerin") Rbasal increased due to addition of organic C and nutrient (Lambais and Carmo, 2008) in both soils. The K inorganic sources (KCl and K 2 SO 4 ) did not stimulate the soil microbes and the "K glycerin" doses were not sufficient to promote higher activity regarding the K mineral sources in the TH soil because the increased soil microbial activity depend on carbon available soil (Araújo and Monteiro, 2006). In soybean harvest (E2), TQ soil only increased the Rbasal with increasing doses of "K glycerin" that differed from K inorganic sources (Table 4). The moisture stress occurred during the experimental period in the cultivation of soybean decreased microbial activity in TH. The higher doses of "K glycerin" apply the TQ may not have been completely decomposed, with a residual effect of C organic in soil. The organic matter added to the soil in the form of organic waste according to the degree of decomposition, may have an immediate or residual soil by means of a slower process of decomposition (Santos et al., 2001). In soybean emergence (E1), the soil microbial biomass carbonic (Cmic) in TH were reduced quadratic "K glycerin" doses (Cmic: ŷ = 348.94 -5.30x + 0.027x 2 , R 2 = 0.90) that differed from K inorganic sources (Table 4), may have been due to an effect of "K glycerin" that harmed the soil microbes. At harvest of soybean (E2) in TH, due to lack of rainfall during the experimental period, the Cmic values were lower at the time E1, with a quadratic decrease with increasing doses of "K glycerin" (Cmic: ŷ = 388.71 -4.30x + 0.017x 2 , R 2 = 0.99), which differ from the inorganic sources of K (Table 4). In TQ, the behavior of the Cmic was different, showing that "K glycerin" stimulated the growth of microbes quadratically in both evaluation periods E1 (Cmic: ŷ = 145.86 + 2.78x -0.012x 2 , R 2 = 0.99) and E2 (Cmic: ŷ = 216.42 + 0.83x -0.003x 2 , R 2 = 0.95), but the K inorganic sources (KCl and K 2 SO 4 ) caused some effect detrimental microbial growth, with reduced Cmic (Table 4). Environmental factors among them such as soil moisture, can modify the ecology, population dynamics and soil microbial activity due to modification of the microbial habitat (Nannipieri et al., 2003).
The metabolic quotient (qCO 2 ) valuated in soybean emergence (E1) may verify that the application of "K glycerin" doses and K inorganic sources caused more stress in both soils (Table 4). Already at the time of harvest of soybean (E2), the evaluation can verify the stabilizing trend in both soils due to exhaustion of nutrients, thus achieving a balance in the middle, by reducing stress soil (Moreno et al., 1999). The application "K glycerin" caused no major changes in the soil chemistry unless the concentration of K and S is a component of this organic waste arising from the production process of biodiesel. Microbiological analysis of soil showed a higher when stress was applied and reducing the end of the experimental period. The values of qCO 2 indicate that the use of "K glycerin" does not cause stress in the soil, thus can be used without major environmental consequences related to their direct application to the soil.
The use of "K glycerin" in the field of soybean yield in soils with average K content, resulting in the recommendation of 87 kg ha -1 of K 2 O to achieve maximum grain yield would result in the removal of 3.5 m 3 ha -1 crude glycerin in the biodiesel industry stocks. The area planted with soybeans in Brazil in the harvest of 2012/2013 was just over 28 million ha (Conab, 2013); thus 1% of this area was used to "K glycerin" would remove national stocks a total of 98 million m 3 of glycerin, making this waste not disposed to harming the environment. The "K glycerin" does not cause environmental problems to the place where it was applied as recommended by the available K in soil, however, studies are needed to evaluate its long-term use to understand the dynamics of the crop yield and soil.