PTQ Q4 2023 Issue

PO hydrotreatment in laboratory-scale units POs can be derived from different raw materials, such as plastic waste or biomasses. They are obtained by process- ing the raw material at high temperatures, either in the presence or the absence of oxygen. The obtained pyroly- sis oil can have a wide range of preliminary contaminants, such as ashes or residual coke, that will be removed in a pretreatment plant before being processed by refineries or pilot plants. The PO has a high variability in chemical composition and properties that depend on the type of pyrolysis and the starting material chosen to produce it. The pH can be as low as 1.5, and water and oxygen con- tent can be up to 30 and 50 wt%, respectively. S and N can be present in refractory molecules, but their processing should be easy to achieve with the classic hydrotreatment catalysts. Many possible metal contaminants can be found (V, Sb, Si, As, Pb, Se, Te) depending on the types of addi - tives used in the initial plastic, for example. Furthermore, significant concentrations of Cl and Br can be found in pyrolysis oil derived from biomasses. However, the main problem of pyrolysis oil is the high reactivity of the feed. Such feedstock is rich in olefins and aromatics that tend to quickly react to polymeric products and clog reactors or even the transfer lines to the reactors. Therefore, it is crucial to fully characterise the feedstock before starting any kind of testing with this type of mate - rial. The metal types in the feedstock can be detected and quantified by XRF. Elemental analysis is employed to iden - tify the O content, while N is quantified by S/N analysis. The TAN is also very important to consider when deciding what type of pipes can be used to perform the tests. The aromatic content and the iodine number are also essential analyses to understand how to correctly process the pyrol- ysis oil. With every new pyrolysis oil to be tested, it is good prac- tice to perform a simple pumping test before feeding it into a running unit. This is meant to ensure that the tubing and pumping system to be used in further testing is suitable for the task. The aim should be to achieve operation without a considerable pressure drop build-up over several days. Additionally, before flowing the feed through the unit, it is recommended to leave a small amount of feedstock at an elevated temperature in a vial to evaluate its temperature stability and understand how much preprocessing is neces- sary. The temperature stability test is especially important if the pyrolysis oil requires a high temperature to be pumped. If, for example, the pyrolysis oil has quite a high iodine number and metal contaminants, it may be possible to use a set-up with three catalysts in series, as shown in Figure 5 . In this case, it would be possible to evaluate the activity of the hydrodemetallisation (HDM) catalyst, the hydrotreat - ment catalyst, and the eventual hydroisomerisation/hydro - cracking catalysts that follow. The HDM and HT catalysts in the first reactor would have to work at very low tempera - tures to avoid any type of polymerisation and subsequent clogging. If the pyrolysis oil is very reactive but with a low metal content, it is recommended to use an already pro- cessed feed to dilute it or grade the activity of the catalyst in

HDM/HT

HI/HC

a full HDO, some remaining monoglycerides, diglycerides, and triglycerides, along with FFAs and alcohols, will be present in the mixture. This is illustrated in the top panel of Figure 4 . If a partial conversion is all that can be obtained out of the HDO catalysts, it is recommended to adjust the offline GC analysis method to account for the response fac - tor of the oxygenated compounds. The same results can be achieved by equipping the FID with a Polyarc system that converts all organic compounds to methane, evening out the response factor of oxygenates and hydrocarbons. The oxygen-bearing molecules burn differently from the paraffins in an FID, making the SIMDIST based on the ASTM standard not optimal for quantification. However, the SIMDIST can highlight the presence of unconverted glycer - ides, as shown on the top panel left-chromatogram (cyan) of Figure 4. This kind of SIMDIST is observed when the tem - perature of the reactor is too low. If the temperature is raised to the ideal level for performing HDO, HVO can be produced with minimal residual oxygenates, as shown by the right chromatogram (orange) in the middle panel of Figure 4. Afterwards, HVO is hydrotreated in the HI reactors to produce fuels. HVO undergoes hydroisomerisation and hydrocracking, resulting in a mixture of linear and branched paraffins with varying numbers of carbon atoms. This latter case is represented by the left chromatogram (red) in the middle panel of Figure 4, which also shows an example of pure VO as a reference on the right chromatogram on the top panel (dark blue). Once the fuel mixture is obtained, it is important to also run additional analyses to evaluate fuel properties such as density, cloud point, pour point, freezing point, and octane or cetane number. first reactor could be employed to host guard material to remove the metals (HDM) and a very low-activity catalyst to start removing the most active molecules at low temperature (HT). The second reactor could be filled with an additional pretreatment catalyst to do some more HT, such as HDO/HDS/HDN. The third reactor would host a higher activity catalyst to perform HI and/or HC Figure 5 Configuration of a 24-fold unit with three reactors in series (up to four three-reactor systems in parallel). The

PTQ Q4 2023