The investigations were carried out on cakes of six “00” genotypes of oilseed rape (Brassica napus L.): three line varieties (Bristol, Express and Navajo) and three hybrid varieties (Artus, Baldur, Titan) which were grown at the site of the Faculty of Agriculture in Zagreb during 2011. The samples were first coalesced and then separated in the separator in two procedures, each of them comprising of eight sub-separations, in accordance to the Rulebook on methods of sampling and quality control 137/2004 (Official Gazette, 2004).

Rapeseed was cold pressed in a disc pressing machine (Electrolux T20, Sweden), after extraction of oil from the seed, the raw material was divided in two groups. The first group of cakes consisted of varieties Bristol, Express and Navajo, while the second group consisted of cakes of three hybrids: Artus, Baldur and Titan. After pressing, all variety/hybrid cakes were mixed in equal proportions within a sample. The investigation process started by grinding in a laboratory grinder “IKA MF 10” with an average screen diameter of <0.5 mm. The samples were then pelletized in a laboratory pelletizer (Pellet Press 14-175; Kahl, Germany), after which the analyses of their energy usability were conducted.

Proximate analysis. The characteristics of the samples were determined by proximate analysis according to standard methods: moisture content (CEN/TS 14774-2:2004) in a laboratory oven (INKO ST-40, Croatia); while ash content (CEN/TS 14775:2004), fixed carbon (by difference), and volatile matter (CEN/TS 15148:2005) were determined by using a muffle furnace (Nabertherm GmbH, Nabertherm Controller B170, Germany).

Ultimate analysis. Total carbon, hydrogen, nitrogen and sulphur were determined simultaneously, by method of dry combustion in Vario Macro CHNS analyzer (Elementar Analysensysteme GmbH, Germany), according to the protocols for determination of carbon, hydrogen and nitrogen (HRN EN 15104:2011) and sulphur (HRN EN 15289:2011). Oxygen content was calculated by difference.

Calorimetry. Heating value was determined by the ISO method HRN EN 14918:2010 using an IKA C200 oxygen bomb calorimeter (IKA Analysentechnik GmbH, Heitersheim, Germany). A quantity of 0.5 g of each sample was weighted in a quartz crucible and put in a calorimeter for combustion. Higher heating value was obtained after combustion, by using the IKA C200 software. Heating value is reported in MJ/kg on dry basis.

The analysis of biogas production was carried out in the laboratory biogas plant of the Faculty of Agriculture in Zagreb. The lab plant consists of a water bath, mixer and a stainless steel batch fed reactor, and provides full simulation of the conditions of a large scale industrial plant for biogas production (Fig. 1). The investigation was focused on the possibility of using rapeseed cake as a raw material for biogas production by adding inoculum. The inoculum used for this purpose was the digested residue from the biogas plant Osatina Ivankovo, Croatia. Anaerobic digestion was conducted in mesophilic conditions at temperature of 35°C during a 40-day period. During digestion process, biogas production was measured daily, and biogas composition was analysed in a gas chromatograph CP-3800 (Varian, USA), in accordance with the method HRN EN 6974-4:2008. The biogas production investigation was conducted in triplicate.

Residue from AD and biogas production presents a very suitable material that can be used as bio-fertilizer. Therefore, this investigation also encompassed the analysis of the material content in order to establish whether it meets the requirements set out in the Croatian Regulation on ecological agricultural production (Official Gazette 10/2011). The pH value was determined directly from the samples in a pH-meter with combined electrode; electroconductibility (EC) was determined by use of a conductometer MA5964 with combined electrode, and total nitrogen was determined by the Kjeldahl method. Phosphorus was determined by molybdate-blue method on a UV/VIS spectrophotometer PU 8600, K and Na were determined by flame photometry and all other elements (Ca, Mg, Mn, Zn, Cu, Fe, Pb, Cd, Ni, As, Hg, Co, Cr) by atomic absortion spectroscopy (APHA, 1995). The samples were burned in a microwave oven Ethos D (Milestone, England), in accordance the supplier application notes for microwave digestion.

Since the design of the experiment was not specifically defined (the allocation was made on an absolutely random basis with various numbers of observations), the average values for each investigated factor were calculated and interaction was included in the relevant analysis. All measurements were carried out in triplicates. All data were analysed according to the generalized linear model (GLM) procedure in the statistical analysis system package version 8.00 (SAS, 1997). Data were expressed as means and, if justified by the statistical probability (p<0.05), subjected to the Duncan’s new multiple range test. Differences were considered as statistically significant if p<0.05.

The proximate, ultimate and physical analyses of the investigated groups of rapeseed cake after pelletizing are presented in Tables 1, 2 and 3. The samples were compared against CEN/TS 14961:2005 as the referent standard for solid biofuels quality.

Table 1 shows that moisture content was approximately (7.00-7.31%) in both samples, which is similar to that of olive cake, as described in the available literature (Al-Widyan et al., 2006; Mirande et al., 2007; Chouchene et al., 2012). Moreover, ash content was somewhat above 7.0% which is a value higher than expected, but it remains within the upper limit of the CEN/TS 14961:2005 standard; high content of ash could be an effect of soil fertilization during rapeseed cultivation. The found fixed carbon values were between 16.52% and 16.84%, while volatile matter was around 69%. Similar results were obtained in the investigations conducted on olive cake by Al-Widyan et al. (2006), Mirande et al. (2007) and Chouchene et al. (2012), where fixed carbon was between 9.94 and 23.4%, while volatile matter was between 67.7 and 80.19%. High volatile matter content indicates that during combustion most of the biomass material will volatilize and burn as gas within the system. On the other hand, high fixed carbon content is a characteristic of herbaceous agricultural biomass residues (Vassilev et al., 2010); it produces char and burns as a solid material in the combustion system (Kreil & Broekema, 2010). Thus, such high levels have positive effect on combustion properties.

The ultimate analysis results show values of C, S, H, O and N content respectively, which are in accordance with the CEN/TS 14961:2005 standard for solid biofuels. Namely, this standard does not set up specific values for rapeseed cake, but does give the ones for olive cake, as a by-product from food processing. Of all elements, only H and N slightly exceed the standard values; this could be due to a more intensive soil fertilization of rapeseed comparing to olive cultivation.

The physical values prescribed in CEN/TS 14961:2005 for wooden pellets are the following: diameter (D) up to 25 mm, length (L) up to 5·D, abrasion (<2.3%) and density (>1.12 g/dm³). Thus, it can be determined that both groups of the observed samples fully comply with the standards for these parameters. CEN/TS 14961:2005 is a norm which sets out energy standard for olive cake among other raw materials that are potential source for solid biofuels production, while standards for rapeseed cake have not been defined. The analysed samples were compared with olive cake because it is the most similar material to rapeseed cake in terms of utilization and production technology. The data of this Technical Specification were obtained from a combination of researches mainly from the Austrian, Dutch, Italian, Greek and Spanish sources. These values describe properties that can be expected in Europe in general.

Heat content, a very important factor which affects utilization of any material as a fuel, is influenced by the proportion of combustible organic components contained in it (Kataki & Konwer, 2001). It is related to the oxidation state of the natural fuels in which carbon atoms generally dominate and overshadow small variations of hydrogen content (Demirbas, 2004). Higher heating value (HHV), defined as a latent heat of the water vapour products of combustion, was found to be above 18.00 MJ/kg, which is in accordance with the applicable standard for solid biofuels, while lower heating value (LHV) was above 16.00 MJ/kg.

In all parameters presented in the Table 1 (proximate analysis), Table 2 (ultimate analysis) and Table 3 (physical and calorimetry analysis) no significant differences were found.

Biogas productivity was obtained for both types of the rapeseed cake samples, and the results are shown in Fig. 2. The rise of productivity was exponential during the first twenty-odd days and the maximum value for these materials (variety cake/ hybrid cake) was reached around 22^nd day of the process. Also, productivity was continuous for about 36 days from the start of the process, when a slight decline followed. Comparing the analysed production of biogas on a daily basis with the data obtained by Tekin & Dalgiç (2000) and Petersson et al. (2007) during AD of olive cake and rapeseed cake, the traceability of the data was observed. The mentioned authors obtained the highest production of biogas on the 20^th day of the process. The production results published by Tekin & Dalgiç (2000) are fully consistent with the data quoted in this paper. This is different from Petersson et al. (2007) who obtained significantly lower production due to the acclimatization of microorganisms.

Table 4 shows medium values of physical-chemical analyses of digested residues (rapeseed cake of varieties and hybrids) after AD. In order to achieve optimal decomposition of the substrate in to methane and carbon dioxide, during digestion, the tendency should be towards neutral value of the digested material. One of the most important parameters for efficient and good quality production of methanogenic bacteria is the pH value. Liu et al. (2008) quote the optimal pH ratio for the most efficient gas production (6.5-7.5). Also, constant monitoring of pH value in digester is essential because its sudden drop may result in premature ceasing of biogas production. However, a reduced pH value (reaction) of the investigated raw materials may be caused by a higher concentration of lignocellulose material during AD, which was corroborated by Al-Masri (2001) in his investigations of digested residue from olive cake. In the investigated samples a mild alkaline reaction of digested residues was found (pH=5.27–5.30), which is most probably caused partly by increased Ca content. It can be concluded that the pH value is approximately within acceptable limits. Determination of electroconductivity (EC) in digested residues is aimed at establishing total quantity of salt in the solution because of soil salinization caused by fertilizers (Voća et al., 2005). The EC values of the hybrid cake/ variety cake were within a range of 2.22 to 2.24 mS/cm. Al-Masri (2001) determined EC of olive pomace at 2.19-2.26 mS/cm and emphasized that EC declines in relation to percentage of lignocellulose substance in the digester as is the case with pH value. It is essential to monitor the content of these key biogenic elements in the substrates, i.e., content of Ca, Mg and Na. The results for Ca content in the digested residues were within the limits of 0.42% for variety rapeseed cake and 0.44% for hybrid rapeseed cake. Contents of Mg (0.37%) and Na (0.04-0.05%) were also determined in the digested residues, and these values were identical in both investigated types of cake. The comparison of the found Ca, Mg and Na values against literature data for olive cake (0.51-10.22%; 0.09-0.67%; 0.02-0.13% respectively) shows a slight variation of Ca percentage content, while Mg and Na are in line with the literature data (Environmental Impact Assessment, 2004). High moisture content usually facilitates AD, but it is difficult to maintain the same availability of water throughout the digestion cycle (Hernandez-Berriel et al., 2008). Also, the chemical analysis of the digested residues showed equal moisture content. So, hybrid cake with 80.40% humidity content had only slightly higher percentage of moisture compared to 80.07% humidity content of variety cakes. It has been reported that the highest methane production rates occur at the humidity level of 60–80% (Bouallagui et al., 2003). Therefore it can be concluded that the investigated substrates have moisture levels almost optimal for biogas production. The analysis of ash showed the content of organic and mineral matter in the samples; ash percentage content was found in a range from 6.88% in digested residue from variety cake to 6.92% in digested residue from hybrid cake, while ash content of 7%, found in olive pomace, is consistent with literature sources (Chouchene et al., 2012).

In order to determine whether the samples of the investigated digested residues are suitable for fertilization of agricultural soil, the ratios were calculated of main biogenic elements -N, P and K. The result was 8.32:1:1.94 for hybrid cake and 8.88:1:2.10 for variety cake. For example, specific ratio of nutrients for cereal grains is 1.2:1:1.5; for potato 1:1:1.8; for grass 2.4:1:1 (Kaltwasser, 1980). It is evident that with such relations the digested residues can meet the plants’ needs for all elements that were analysed and that there is an increased level of nitrogen, as in the case of pellets. In general, bio-fertilizers have a narrow C/N ratio, from 10:1 to 15:1 (Voća et al., 2005). The same C/N ratio was found (8.7:1) in the investigated digested residues of rapeseed cake. Also, the ratio observed was somewhat lower than the one previously quoted.

Determination of heavy metals in the digested residues is important not only for environmental considerations but also because of the methanogenic bacteria in AD. For efficient growth of all methanogenic bacteria relatively high levels of iron, nickel and cobalt are required (Ram et al., 2000). However, optimal concentrations of iron, nickel and cobalt were found in the digested residues in very wide variations (Gonzales-Gill et al., 1993; Jarvis et al., 1997; Mochinaga et al., 1997; Takashima & Speece, 1997; Basiliko & Yavitt, 2001; Kida et al., 2001). Such variations can be explained by presence of different varieties of methanogenic bacteria in the substrates and their different and particular needs for iron and cobalt. Due to this, low levels of some of these metals may result in limiting the whole process of biogas production. On the other hand, a higher metal concentration may cause toxicity, i.e., it may prevent development of methane producing bacteria (Nies, 1999). In terms of environmental protection, damaging substance is any substance which is found in the agricultural soil in such concentrations which either temporarily or permanently compromise its fundamental role of being appropriate habitat for cultivated and feral vegetation. Allowed is the utilization of digested residues on arable surfaces, meadows and plain pasture grounds where soil contains some of heavy metals and persistent organic harmful matters in amounts below 50% of limit values set out in the regulation on ecological agricultural production in the Republic of Croatia (Official Gazette 91/2001). As it is evident from the data shown in Table 5, the concentrations of these heavy metals found in all samples of digested residues are below the prescribed limits and meet the requirements set out in the Republic of Croatia and, therefore, can be used without restrictions in ecological agricultural production.

Similarly to the parameters found in proximate analysis, ultimate analysis and physical and calorimetry analysis of rapeseed cake, chemical analysis (Table 4) and heavy metal analysis of the digested residue (Table 5) showed no significant differences.

With the production of the first generation biodiesel fuel in the Republic of Croatia as well as with the production needs in the future, Croatia is faced with the excess of rapeseed cake as its by-product. As an effect of the Homeland War the Croatian livestock has been reduced by 50% from the previous levels, which seriously limits the possibilities of the livestock farming to absorb total amounts of rapeseed cake. For these reasons, this work investigates the potential utilizations of rapeseed cake as an energy product, in form of solid fuel pellets as a raw material for production of biogas by alkaline digestion in biogas plants. Since this process leaves digested residue, this investigation aims to assess the quality of this by-product. In addition to variety rapeseed, there is an increasing use of hybrids in the Republic of Croatia, so the investigations included the cakes of rapeseed “00” cultivar varieties (Bristol, Express and Navajo) and three hybrids (Artus, Baldur, Titan) in order to identify eventual differences. Solid fuel from variety/hybrid rapeseed cakes was produced by pelletizing because of higher quality of the product i.e., its better physical-chemical characteristics. The results showed no significant differences between the variety/hybrid cakes when the pellet production is concerned. Namely, physical-chemical analysis of the pellets discovered approximately equal parameters (radius, length, moisture, density and abrasion). However, it can be noted that there is a somewhat higher content of ash as well as an increased level of fixed carbon and volatile matter, which is most probably a characteristic of herbaceous agricultural biomass residues. The results of the ultimate analysis show contents of C, S, H, O and N, which are in accordance with the CEN/TS standard for solid biofuels. Namely, this standard does not set up specific values for rapeseed cake, but does give the ones for olive cake, as a by-product of food processing. Of all elements, only hydrogen and nitrogen contents slightly exceed the standard values. Higher heating value (HHV), defined as latent heat of water vapour products of combustion, was found at levels above 18.00 MJ/kg, which is in accordance with the applicable standard for solid biofuels, while lower heating value (LHV) was above 16.00 MJ/kg. It results that 1 kg of oil (41.00 MJ/kg) can be replaced with 2.56 kg of pellets obtained from the cakes of variety/hybrid rapeseed.

In addition to investigating pelletizing cake by thermal processing, the aim of this investigation was to determine the possibility of using variety/hybrid rapeseed cakes in biogas production as well as the quality of digested residue from biogas production. When the biogas production from variety rapeseed cakes is compared to that from cake of hybrid origin, there is no significant difference between the two materials. However, the analysis of biogas production shows that maximum production is obtain on around 20^th day of the process, which is relatively late. It is proposed that the rapeseed cake be used only as one of the components in a mix with other ingredients, rather than as a single component. After completion of the biogas production process it can be determined that there are no significant differences in the chemical composition of the digested residue between cakes of variety origin and those of hybrid origin. Additionally, the values of all the elements required under the Regulation on ecological agricultural production in the Republic are such that the obtained digested residue is assessed as suitable for use in the agricultural production as fuel.