Nov-2023

Catalyst testing for renewable fuels and chemicals

The possibilities and best practices to run renewable feedstocks in high throughput units are presented.

Giada Innocenti, Kai Dannenbauer, Jochen Berg, Xavier Sanz and Ioan-Teodor Trotus
hte GmbH

Article Summary

Worldwide sustainability goals are calling for a major redesign of the petrochemical and chemical industries. The net-zero CO₂ emission goals set for the next decade require such industries to improve their efficiency and start relying on different raw materials. Selected resources need to be either renewable, such as vegetable oils, animal fats, and sugars, or recycled, such as pyrolysis oil or used cooking oil.

Additionally, they need to be widely available and not food competitive. Food competition makes using most of the feedstocks that are well-characterised and commercially ready (vegetable oils) challenging in the long term. Thus, the need to test feedstocks deemed more challenging, such as pyrolysis oil, will become very relevant soon. The use of such feedstocks will also perfectly match the concepts of a circular economy that aim to use the wastes of one industry as feedstocks for another. However, such types of feedstocks are currently not close to commercial implementation and require additional developmental work.¹

State-of-the-art tools are required to fill the gap in the development and optimisation work required for the efficient conversion of renewable raw materials. High throughput experimentation technology can be employed to test many reactors in parallel under relevant process conditions. Experimental data can be generated at an unprecedented rate, enabling acceleration of process development, the quick choice of the most economically productive catalytic system from many options available, the development of new catalysts, or simply the execution of proof-of-concept studies. This type of testing is very important in fluid catalytic cracking and hydroprocessing to evaluate the performances of different catalytic systems.²

High throughput experimentation enables a quick and effective exploration of the experimental space. It can be easily associated with design of experiment (DOE) to get the most out of a limited number of tests.

Regardless of the feedstock and the experimental plan to be run on a high throughput unit, preparation and validation are key. It is important to make sure that liquid, gas, and H₂ recovery are 100% to ensure high data quality and reproducibility.

The renewable resources discussed here are vegetable oils (VO) and pyrolysis oil (PO). These two feedstocks are at very different technological development stages, as they present different challenges associated with their use.¹

VO hydrotreatment in laboratory-scale units
VOs are, in general, mixtures of triglycerides. A triglyceride molecule is composed of three fatty acids esterified with a molecule of glycerol. Fatty acids have an even number of carbon atoms, most often arranged as chains between 16 and 20 atoms long, and a variable number of unsaturation (typically 1 to 3 double bonds). VOs contain up to 15% O, which can be removed by hydrodeoxygenation (HDO) or decarbonylation (DCN)/decarboxylation (DCX) (see Figure 1).

The S and N content is usually very low (up to 30 ppm). Contaminants such as P and Si, whose concentration depends on feedstock derivation, can cause catalyst deactivation and even reactor clogging if present in high amounts. Free fatty acids (FFA) and Cl in the feedstock can also cause corrosion. It is, therefore, important to know the composition of trace elements in the feedstock before starting a pilot plant testing, as well as the total acid number (TAN) and the bromine or iodine number (unsaturation degree).

Standard hydroprocessing technologies can be applied to convert VOs to fuels if they can be tuned to overcome the new challenges posed by the different chemical make-up of renewable feedstocks. These challenges include higher O content, corrosion issues, and different types of impurities. High throughput units (16- or 24-fold) can host reactors with an inner diameter of up to 8 mm, while bench-scale reactors (4- or 8-fold) can be used with reactors up to 19 mm of inner diameter. The bench-scale units also allow testing guard bed material to eventually process feedstocks with high content of P or Si and evaluate materials based on their performance for P and Si uptake. If the iodine number of the feedstock is high and clogging of the reactors is experienced, it is possible to optimise the reactor loading by exchanging the top layer of inert with a lower activity catalyst that will hydrogenate the unsaturation.

To be converted to fuels, VOs need to be hydrodeoxygenated and then dewaxed (hydroisomerisated) to convert the linear paraffins into branched ones. The first step in the hydrotreatment of VOs is the hydrogenation of all the unsaturation, as shown on the top panel of Figure 1. The FFAs are then undergoing HDO, DCN, or DCX to yield linear paraffins. However, the last three reactions mentioned may also take place on the hydrogenated triglyceride or any di- or monoglyceride still present in the feed. HDO is the preferred and targeted reaction since the co-product is H₂O, while DCN and DCX are unwanted reactions that co-produce CO and CO₂.

All the co-products can cause possible issues downstream of the reactor. H₂O together with HCl, formed from trace amounts of Cl in the feed, can promote corrosion. The possible corrosion issues raised by the presence of water in the liquid phase can be solved by keeping as much water as possible in the gas phase. Vaporising the water is also helpful for the oxygen balance that can be affected by the separation efficiency. The H₂O and organic phases can be separated only to the extent they did not emulsify, and the water is not micro-dispersed. When it is not possible to keep all the H₂O in the vapour phase, it is good practice to run an elemental analysis and/or Karl-Fischer titration of the organic sample to account for dispersed/emulsified water.

The oxygen balance can be significantly improved, as shown in Figure 2, by increasing all temperatures downstream from the reactor. Additionally, it is recommended to introduce some inert gas after the reactor to dilute the concentration of H₂O in the gas phase and promote its evaporation. In a single-stage configuration, H₂O, CO, and CO₂ will enter the hydroisomerisation (HI) reactor, possibly deactivating the noble metal catalyst (left panel in Figure 3). The sour mode of operation should only be chosen if the HI catalyst can work in such conditions or if its resilience to such contaminants is under evaluation.

If the water effect on the HI catalyst performance is under evaluation, the test should be run at a specific temperature that yields a product with a specific target parameter value (such as density, cloud point, or pour point). The target parameter must be routinely tested to ensure that the system is operating correctly. Afterwards, once the behaviour of the system is well characterised, it is possible to increase the concentration of H₂O in the feed by introducing a larger amount of water to the gas phase and creating a system upset that can eventually be exaggerated by increasing the temperature by 5-10°C.