The topic of this paper deals with biomass briquettes from long hemp, reed and reed canary grass stalks. The orientation of straw or reed stalks had to promote binding by the pressing operation. The specific briquetting energy of coarse chopped arranged reed and hemp stalk particles ranges from 51.61 kJ·kg -1 to 67.23 kJ·kg -1 . Arranged structure of biomass particles in briquetting die is recommended for significant increasing of durability of stalk material briquettes. The splitting force of the hemp stalk briquettes ranges from 110 N·mm -1 to 122.37 N·mm -1 , the splitting force of the arranged reed particles lies between 65 N·mm -1 and 80 N·mm -1 . The specific splitting force is 2 to 5 times higher than for non arranged stalk briquettes. New briquetting equipment is necessary to be designed for biomass particle arranging before pressing. The objective of the study is to develop a method for determination of the specific energy for stalk materials. The article presents the specific energy calculation method and the results of experiments for determining the specific energy for different stalk biomass. The measurement results show that the specific energy of hemp stems reaches 274 J·m·kg -1 , 50.3 J·m·kg -1 for lake reeds, and for reed canary grass - only 7.8 J·m·kg -1 . These values will be used in calculation of the energy consumed in the cutter. briquette density � = 0.8-1.2 g·cm -3 . The density of � > 1.0 g·cm-3 is recommendable for high quality wood briquettes. This value had been used for evaluation of herbaceous material densification results. In the previous investigations it was stated that the density of the arranged reed and hemp stalk particles exceeds the recommended in the standards 1000 kg·m -3 and reaches the value 1185 kg·m -3 for the arranged hemp stalk particles with the length 150 mm and briquetting pressure 212 MPa. The splitting force of the hemp stalk briquettes ranges from 110 N·mm -1 to 122 N·mm -1 , the splitting force of the arranged reed particles lies between 65 N·mm -1 and 80 N·mm -1 (1). New briquetting equipment is necessary to be designed for biomass particle arranging before pressing. Reeds in several countries are commonly used for roof building. The necessary material for it is collected in bundles. This harvesting technology could be used also for reed material planed for briquetting. Stalks in bundles already are oriented in the direction necessary for it. Actuality of this paper is connected with the bundle cutter which provides stalk orientation after cutting. Crushing energy is an important parameter for characterizing biomass machines. Crushing energy reduction reduces the cost of briquettes and increases the competitiveness of the market for that fuel. The objective of the study is to develop a method for determination of the specific energy for stalk materials. Materials and methods The study was carried out in the Institute of Mechanics, Faculty of Engineering. There are several methods used in biomass cutting. It has been proved that the lowest energy consumption of stalk shredding provides shear cutting (2). In the result of the research a method with two rotating knife rollers was developed, Fig. 1. The experimental model of the cutter was designed and experimentally tested, Fig. 2. The cutter consists of two rotors with blades that rotate with the synchronization gears 3. The blades 2 are mounted on the rotor and are powered by an electric motor 4. Biomass (common reed stalks) is fed to the knives on the tray 5. Angular velocity of the rotor was measured using the tachogenerator 6. In previous studies geometric and kinematic parameters of the stalk biomass cutter were evaluated (3). To determine the energy of the biomass stalks, the experimental tests were carried out. Energy consumption for stalk has been investigated using experimental equipment, Fig. 2. To calculate the energy the electric drive voltage and current consumption were measured. The measurement data were recorded with a 12-bit virtual data logger Picoscope 4424. Measurement error for this data logger does not exceed ± 1 %. Further the obtained data were imported in Excel