https://store-images.s-microsoft.com/image/apps.62890.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.105cb921-9945-4ab4-983f-c96bc6d17d19.5ae94135-07cb-4c53-9149-39bdfb72f440

Process Operations Advisor

by Beyond Limits

AI suite to increase oil refinery profitability and lubrication formulation times

REFINING SITUATION

As a petroleum refiner, you need to increase operational efficiency to realize tight-margin targets.  Profitability in a refinery is always at risk, margins are tight, decision-making is complex, and variability is a challenge to current processes. Achieving optimization goals requires the uniform application of best practices and unique human expertise to achieve planning objectives.


SOLUTION

Beyond Limits refinery operations advisor is a web-based software system that empowers refinery operations teams to plan, operate, and improve process and product manufacturing efficiency and profitability. The system is designed to “think” like an engineer and provide expert guidance to solve problems across process units. It goes beyond conventional asset-only focused approaches to ensure process objectives are met.

Our AI technology supports the human decision loop similar to the way in which Advanced Process Control (APC) supports the control loop. As a result, our customers have estimated an increase of heavy feedstock utilization by up to 10%.

LUBRICATIONS FORMULATION SITUATION

The process of creating, optimizing, and testing new engine lubricants has been a long, high-risk, high cost, trial and error process that requires deep domain expertise. In addition, ever- changing regulatory, environmental, and equipment manufacturers’ guidelines further complicate the process. Lastly, each candidate formulation is typically validated through a long and expensive series of tests, further increasing the cost, risk of failure, and time- to-market for new products.

SOLUTION

Beyond Limits’ Cognitive Formulation Advisor leverages your historical data (historical blend composition and measured properties) to propose new hypothetical blends.  Through this process, Beyond Limits incorporates your formulation domain knowledge, alongside parameters and directional guidance (such as regional availability of base stock) into the AI-driven solution.

The result is a well-integrated software product that recommends hundreds of viable blends meeting constrained viscometric targets with high confidence in predicted properties in a fraction of time compared to the alternative approaches. The solution also explicitly characterizes the confidence level in its prediction and recommends a series of optimized blends by considering cost and performance. A few select lead candidates can be created and tested with results verified in your laboratories. Customer results suggest that Beyond Limits' solution was more than 95% accurate on average.

At a glance

https://store-images.s-microsoft.com/image/apps.9659.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.be11aa11-9f26-4668-8db9-6d627717de23.c72f37f9-86a8-45e3-8a32-b8f939b1ae2e
/staticstorage/linux/20241223.2/assets/videoOverlay_7299e00c2e43a32cf9fa.png
https://store-images.s-microsoft.com/image/apps.40561.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.0dd702ec-6722-40ea-be13-acdcdd229d27.76df9920-e920-414c-9cbf-04cfe0a264a7
/staticstorage/linux/20241223.2/assets/videoOverlay_7299e00c2e43a32cf9fa.png
https://store-images.s-microsoft.com/image/apps.33742.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.aac88aa1-8548-4044-b60d-6b33dbe0a11a.43ab2241-19ec-4876-a8b5-f78e6f527551
https://store-images.s-microsoft.com/image/apps.15377.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.aac88aa1-8548-4044-b60d-6b33dbe0a11a.f24710a5-6b13-4cc1-adaa-9f14fc217170
https://store-images.s-microsoft.com/image/apps.15276.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.aac88aa1-8548-4044-b60d-6b33dbe0a11a.05601d07-41b8-47b2-94ec-c70bbf80b476
https://store-images.s-microsoft.com/image/apps.53200.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.aac88aa1-8548-4044-b60d-6b33dbe0a11a.cefa1626-3a1f-48cd-8989-4d82fd133d67
https://store-images.s-microsoft.com/image/apps.56930.1ebc1587-3a2f-4f16-82cc-c9e0e2b19195.aac88aa1-8548-4044-b60d-6b33dbe0a11a.82734ae0-1ca1-4ed4-9d38-193cb006a977