UK-based Modcon Systems Ltd has drawn attention to a recently published patented method for fuel blending that addresses one of the most persistent technical problems in refinery operation: how to maintain product quality and cost performance when component streams are constantly changing under live plant conditions. Modcon Systems Ltd is listed on Companies House and describes its work as focused on process analysis and AI-enabled optimisation for process industries.
The patented method was filed in May 2021 and first published as an application in December 2022. The B1 publication appeared on 1 January 2026. The patent is titled Fuel blending and lists Gregory Shahnovsky, Ariel Kigel and Ori Honigman among the original assignees and inventorship record.
Although in-line blending is a long-established refinery practice, the patent description argues that the process remains highly dynamic because the chemical composition and physical properties of incoming process streams can shift continuously. It notes that gasoline, diesel and other petroleum products are built from multiple refinery streams and that profitable blending requires more than simple ratio control. Production capacity, intermediate tank capacity, component cost, changing product grades and the need to minimise giveaway all have to be managed at the same time.
According to the patent description, the core of the method is an integrated blending system that uses at least one process analyser to assess parameters in different process streams, compare blended-stream quality against predetermined criteria and pass that information to a blend optimiser. The optimiser then predicts new parameters for the different streams so that the control system can regulate the flow rate or amount of each stream in response. The patent also describes optional use of a neural network, historical data storage, flow sensing, blend quantity sensing and a static mixer as part of the wider arrangement.
This is technically significant because conventional blending control often struggles with the gap between real process behaviour and the slower pace of trusted quality confirmation. The patent description points out that traditional ASTM-based analysers such as RVP, distillation, octane and cetane analysers can offer accurate measurement but often respond too slowly for rapid corrective action. It also notes that correlative spectroscopy systems such as NIR and FTIR provide faster multi-property measurement but can drift when stream chemistry changes in ways that affect spectral response without matching changes in physical properties.
The method is therefore framed not simply as a new blending recipe tool but as a way to connect live measurement, quality comparison, prediction and repeated adjustment into one control structure. The patent states that the process can repeat the analysis and prediction loop until the measured quality of the blended stream substantially matches the target criteria. It also says that predicting the new stream parameters can use linear programming and or neural-network-based dynamic modelling.
A further technical point in the patent is its emphasis on the wider production chain. The description says that optimising individual refinery units does not guarantee optimum performance for the final blending station because the operating mode of upstream units affects the quality and economics of the finished product. It argues for focusing on key performance indicators directly linked to physical and chemical properties and for using those indicators to predict stream behaviour and adjust fuel production set points across the chain.
For the refining sector, that message is timely. Fuel blending is often treated as a mature discipline, yet the published patent underlines that major practical challenges remain at the interface between real-time analysis, process control and plant-wide optimisation. In mature engineering fields, new protected methods tend to matter most when they address operational gaps that have resisted standard solutions for years. This patent appears to sit squarely in that category.
Modcon says its broader focus includes process analysers, analyser systems and optimisation technologies for sectors including oil refining, natural gas, pipelines, chemical and petrochemical industries. The newly published fuel blending method fits within that wider industrial context and adds to ongoing industry discussion about how best to combine live data with practical control action in complex production environments.
