In anticipation of Beyond OPTIMIZE: New Insights in Performance Engineering and Production Optimization next month, our blog editor spoke with Seçil Uzun, Superintendent of the simulation unit at Tüpraş to discuss her presentation: “Refinery Increases Profits Through Improved Planning Accuracy Using AI-based Aspen Hybrid Models™ and Aspen HYSYS®.”
Marilyn Guisbond: Welcome Seçil; thank you for taking the time to discuss your session for Beyond OPTIMIZE. Can you tell us about your current position at Tüpraş: how long have you been there and what prior industry experience you have to share?
Seçil Uzun: I am the Simulation Superintendent at Tüpraş headquarters, where I have worked for 10 years. In this position, and as a member of the simulation team, I am in charge of creating simulation models of all relevant refinery processes, in addition to executing optimization studies, pinch analysis and revamp projects. This year we are focused on digitalization projects like digital twins and hybrid models. My prior industry experience was in crude and vacuum distillation units as an operations and process engineer. I also worked on several phases of Tüpraş’ Fuel Oil Residue Upgrade Project (RUP).
MG: Can you tell us where Tüpraş is on its digitalization journey, particularly as it impacts your team?
SU: Tüpraş implements various innovative operations in parallel with the Industry 4.0 principles such as online tracking systems, wearable technologies, field implementation of wireless and mobile solutions, robotic technologies and artificial intelligence applications. We prioritize safety with a “human-first” approach in all our digitalization efforts.
Our team has successfully implemented projects on AI-driven hybrid models and presented these studies to the wider organization, including heat exchanger fouling analysis. We will discuss the studies we’ve done on reduced order hybrid models during our session at Beyond OPTIMIZE. In addition to hybrid models, we are currently implementing digital twin projects for CDU optimization and preheat train fouling, adding value in our digitalization journey.
MG: Do you think your digitalization planning has been impacted by the pandemic?
SU: Yes, it is a bigger initiative than I initially expected. Simulations have increased exponentially this year.
MG: Let’s discuss your session at Beyond OPTIMIZE on Aspen Hybrid models and Aspen HYSYS…Can you provide a couple of highlights from your presentation: "Refinery Increases Profits Through Improved Planning Accuracy Using AI-based Aspen Hybrid Models™ and Aspen HYSYS®”?
SU: Updating planning models in Aspen PIMS-AO® using Aspen HYSYS reactor models, aka the “conventional way” provides increased profitability. In our session, we will discuss how profits can be further increased with the application of the recently released Aspen Hybrid Models. Our project provided a sustainable way to improve profitability by using the Aspen Hybrid Models approach, as it takes into consideration thousands of alternative operating conditions, such as feedstock alternatives in relatively quick time. Saving time in updating planning models is an outstanding benefit of Aspen Hybrid Models.
MG: How does Aspen HYSYS fit into this picture?
SU: Aspen HYSYS is fundamental to both approaches: the “conventional way” and the new Aspen Hybrid Models approach. In the conventional way, we were only able to simulate a maximum of 10 operational alternatives, to update the planning model. This approach involved cumbersome workflows that included working with many spreadsheets. Using Aspen Hybrid Models, we have been able to increase our ability to run simulations of operational alternatives from a maximum of 10 to 1,000 – 2,000 in relatively quick time. We think that the broader range of operational scenarios analyzed significantly reduces the number of model revisions throughout the year. Tens of cases vs. thousands has made a huge difference, as well as saving time, especially for our planning team. What used to take 2-3 weeks for my planning colleagues, now takes considerably less time, which impacts profitability, as well as the value of time saved.
MG: Profitability is a key metric in digitalization planning – can you give us an example of how Tüpraş has increased profitability using Aspen Hybrid Models and Aspen HYSYS?
SU: Profitability along with continuity of operations is of vital importance to our business. Efficiency and effectiveness of our planning models have been improved through our use of Aspen Hybrid Models by increasing the operational alternatives considered. By doing this, the nonlinearity of the process is captured in the planning models, which results in more accurate and realistic margin estimations. This increased accuracy of planning models allows more realistic plans and higher profitability.
MG: How do you see Aspen Hybrid models helping the existing workflow in your refinery to update planning models? Would you say you are enabled to do more frequent updates of planning models now?
SU: The Production Planning Department and our simulation team have agreed that Aspen Hybrid Models helps us save time and money by updating planning models. We can do it in much less time, which is a big difference for us. Additionally, as I said before, we think that the broader range of operational scenarios analyzed significantly reduces the number of model revisions throughout the year. We expect that results in the long term will confirm our initial evaluations.
MG: How easy or intuitive was it to adopt Aspen Hybrid models?
SU: Adoption of Aspen Hybrid Models was easier than we expected. AspenTech support has made the adoption process much easier as they were helpful in both guiding us in configuring the models and analyzing the results.
MG: Now that we’ve heard a little about your session, we’d love to know what other sessions you are looking forward to attending?
SU: I am excited that we will have an opportunity to hear from industry experts about their experiences and learn best practices that can create value for us. I am looking forward to listening to the presentations especially on digital twins, hybrid models and carbon capture.
MG: Thank you Seçil. Is there anything else you’d like our audience to know?