Problem:
Petronas’ supply chain struggled with delays, unpredictable oil and gas logistics, tank space issues, and slow planning, making it difficult to manage and costly to fix.
Solution:
The company built a simulation model in AnyLogic to spot bottlenecks, using auto-pulled data in Excel, a web app with maps and charts, daily email updates, and a three-step process to tweak and test plans.
Results:
- Planning sped up from hours to seconds with daily email updates.
- Risks shrank with Monte Carlo tests catching issues like tank tops early.
- Teams synced up better using shared model information.
- Mass balance tracked product amounts clearly.
- Lessons on tricky pickups led to machine learning ideas and vessel tweaks.
Introduction: about Petronas
Petronas is Malaysia’s biggest name in the oil and gas industry, handling everything from refineries to depots and the logistics that connect them across the peninsula and East Malaysia. Beyond its energy operations, Petronas is globally recognized for its long-standing partnership with Formula 1.
This case study looks into how Petronas examined its supply chain, from the bustle of refineries to the far-flung depots, and found a way to manage complexities using oil and gas optimization strategies.
Problem: navigating logistics uncertainties in oil and gas
Petronas’ supply chain is a complex web, stretching from refineries to depots across Malaysia, and it comes with its challenges. At the refinery level, production can encounter difficulties, such as delays or unexpected shifts that disrupt the entire schedule.
Then there’s the oil and gas logistics: vessels moving products often encounter problems, especially with multi-drop routes where they stop at one depot and then another, all while demand swings unpredictably. This makes managing inventory a real challenge.
Depots store products in tanks, some shared with other oil and gas companies, adding another layer of coordination to an already complex situation. Another challenge is “tank tops,” when a vessel shows up too early, and there’s no room to unload, leaving valuable product stuck onboard.
On top of that, planners would spend hours piecing together data by hand, with no clear way to spot risks like these ahead of time. It was difficult to build a solid plan without clear visibility between refineries and depots, and solving problems later often had a significant impact on the budget.
Solution: simulating success in oil and gas optimization
To address these challenges, Petronas developed a simulation model using AnyLogic, designed to identify bottlenecks and validate oil and gas optimization outputs.
The model integrated data from diverse sources such as opening stocks, lifting profiles, vessel routing plans, and refinery production forecasts. It then automatically compiled this data into an Excel file for user-friendly scenario testing.
Running on a monthly horizon, the simulation was paired with an oil and gas optimization model (a mixed integer linear program) to generate initial vessel plans. It then refined these plans by accommodating real-world nonlinearities and uncertainties.
A custom web application, powered by the AnyLogic Cloud API, provided planners with intuitive access. It featured a map view of depots and refineries with highlighted warnings, such as yellow for critical events. Additionally, there was a summary view with mass balance, stack charts, and key performance indicators (KPIs) like tank top incidents.
An automated daily task pulled fresh data each morning, ran the simulation for the next 30 days, and emailed planners a report flagging potential issues. Planners followed a three-step process: run the baseline simulation, adjust inputs (e.g., vessel schedules) to mitigate risks, and execute Monte Carlo experiments to test plan resilience.
Why did Petronas choose AnyLogic?
Petronas chose AnyLogic for its simulation model because it suited the company’s needs. Here are the key reasons:
- Strong Java base: AnyLogic uses Java, which runs fast for big models and performs better than Python, which slowed a past Petronas project. It also has a large community with tools to help build organized setups.
- Helpful Cloud features: AnyLogic Cloud enabled Petronas to grow the model easily, share it quickly with the team, and connect it to web tools and Python for reports.
- Easy to shape: AnyLogic allowed Petronas to adjust the model to fit its supply chain, using ready tools like one for fluids to speed things up.
- Clear visuals: AnyLogic is a tool that enables users to create clean and easy-to-understand visuals.
Are you working in the oil and gas industry and interested in simulation modeling? Learn more from our case studies involving companies worldwide that use AnyLogic. Explore further and discover why it is the leading simulation modeling tool.
Results: paving the way for logistics improvements
The simulation model greatly impacted how Petronas handled its oil and gas logistics, turning a challenging supply chain into something more manageable. It brought clear benefits that helped the company every day and set up new ways to improve in the future. Here are the key outcomes:
- Planning has become faster and easier, going from hours of work to just seconds because data comes in automatically. Planners now get an email every morning with a 30-day view of what’s ahead.
- Risks have gotten smaller with Monte Carlo tests, which show planners problems like tank tops before they hit, saving money by fixing things early instead of after they go wrong.
- Teams work better together now as refineries, depots, and oil and gas logistics employees all see the same information through the model. This allows them to discuss and plan as a group instead of guessing what’s happening.
- New information like mass balance shows precisely how much product is where, helping planners keep track of everything from the refinery to the depot in a way they couldn’t before.
- Petronas discovered that predicting customer pickups at depots is challenging, prompting them to explore machine learning for more accurate forecasts. Additionally, the business is seeking to implement automatic adjustments to vessel schedules and enhance ship repair monitoring.
With AnyLogic, Petronas has made its oil and gas logistics stronger and smoother, and it’s ready to take on more challenges in the oil and gas optimization supply chain in the future.
The case study was presented by Marc Escandell Mari and Nurul Musfirah Murad from Petronas at the AnyLogic Conference 2024.
The slides are available as a PDF.
