scale-up to commercial units. Both empirical and theoretical studies emphasise the impor- tance of maintaining proper reactor geometry, specifically, a high reactor length-to-particle diameter ratio (L/d p ) to minimise axial dispersion and a sufficient reactor-to-particle diameter ratio (D r /d p ) to promote uniform radial liquid distribution. It is possible to mitigate these
Summary of key parameters to control during pilot plant testing
Key parameter
Objective
How to do it?
Minimise axial dispersion 1
Prevent back-mixing
Ensure full catalysts wetting 2
All particles can contribute to the overall reaction/conversion
Mitigate wall effects 3
Uniform temperature and concentration profiles
Table 2
require adjusting the catalyst particle size or incorporating diluent to distribute the flow uniformly. Additionally, catalyst loading techniques (for example sock vs dense loading) can strongly influence local flow behaviour and must be carefully considered. Although hydrocarbon feedstocks tend to promote cata- lyst wetting due to their viscosity, complete wetting does not guarantee effective utilisation. Poor liquid distribution can lead to stagnant zones, inhibit mass transfer, and limit access to active catalyst sites. Maintaining dynamic irrigation, where liquid films are continuously renewed, is vital to preserving catalyst activity, particularly under kinetic-limited regimes.⁶ Industrial-scale units benefit from high superficial veloci - ties, which naturally promote flow uniformity and minimise dispersion. However, pilot and micro-reactors often operate at much lower velocities, making them more prone to axial and radial maldistribution. While inert packing and struc- tured diluents can assist with flow distribution, they do not fully eliminate localised bypassing, especially in small-diam- eter reactors where wall effects dominate. As such, main- taining proper Pe, L/d p , and D r /d p ratios is essential to ensure that pilot test data reflects true catalyst performance.⁶ Figure 4 shows the relationship between L/d p and Pe. The plot includes: • A blue curve showing Pe increasing with L/d p , assuming a constant superficial velocity-to-dispersion ratio. • Dashed turquoise lines at Pe = 40 and Pe = 100, marking thresholds for acceptable and ideal flow behaviour. • A vertical red line at L/d p = 350, the recommended value for achieving plug flow in pilot plant studies. Hydrodynamic and analytical considerations in third-party testing Modern third-party pilot facilities have become significantly advanced in analytical precision, providing continuous data collection and excellent mass balance closure. However, the geometry of high-throughput reactors inherently introduces hydrodynamic limitations. Reactors with catalyst volumes below 10 mL and small-diameter extrudates are more prone to wall effects, axial dispersion, and incomplete wetting. As a result, hydrogen-liquid contact and overall mass trans- fer behaviour can deviate from that in industrial trickle bed systems. While such systems are highly valuable for rapid screen- ing and trend identification, they are not suited for assessing deactivation or stability behaviour, where liquid film thick - ness and hydrogen gradients play a major role. For reliable
potential operating issues in a well-designed pilot plant test, but not in a high-throughput micro-scale reactor due to the extremely small catalyst volumes used in these units. The microreactors are ideal for catalyst screening tests, but not for a proper quantitative catalyst comparison. Due to the pre- viously listed issues, the data from the microreactors often show that the best catalyst is the worst product, and the worst catalyst is the best product, often due to H₂ starvation. The reliability of pilot plant results is highly sensitive to the chosen temperature strategy. When tests are initiated at excessively high inlet temperatures, the product sulphur can quickly drop below detectable levels, making it impossible to distinguish any intrinsic difference in the catalyst activities. To obtain representative results, the test should be started at moderate severity and progressively increased, allowing clear separation between catalysts and avoiding premature deactivation. This approach ensures that the kinetic regime is captured and the true activity ranking between the catalyst brand is maintained. Axial dispersion is quantified by the Peclet number (Pe), which defines the ratio of convective to dispersive trans - port along the reactor bed. Higher Pe values indicate near plug-flow behaviour, which is essential for accurate kinetic measurements. In hydroprocessing systems, a Pe > 100 is typically recommended to limit axial back-mixing and ensure representative catalyst performance. Values between Pe = 40-100 may be acceptable for semi-quantitative testing, but still reflect some deviation from ideal flow. Reactors operat - ing with Pe < 40 exhibit strong dispersion and are generally unsuitable for kinetic evaluation.⁴ Achieving high Pe values requires a sufficiently long reac - tor relative to particle size – L/ d p > 350 is commonly targeted in bench-scale systems.1 This becomes increasingly difficult in micro- and high-throughput reactors due to geometric limitations. In practice, well-designed pilot plants typically achieve Pe values above 100, enabling near plug-flow behaviour suitable for kinetic interpretation. By contrast, micro-reactors, owing to their short bed length and very low linear velocities, often operate in the Pe = 20-40 range (ver - sus a minimum of 100), well within the strong-dispersion regime where mixing effects dominate and true plug flow behaviour cannot be attained. Similarly, maintaining an appropriate D r /d p ratio is critical to avoid wall channelling and radial maldistribution. A minimum of D r /d p > 10 is generally needed to reduce wall effects, while D r /d p > 25 is often recommended for pilot testing with heavy feeds or low liquid velocities.4 Reaching these values may
31
PTQ Q1 2026
www.digitalrefining.com
Powered by FlippingBook