A Data Science Central Community
[Purpose: get Engineers & Scientists thinking outside-the-
box ... think -large- problems. What's possible today vs. needs for tomorrow?]
Oil production dependents on many factors; e.g. Supply, Demand, present inventory, etc. A oil company may have many refineries with many distillation units. How can a company simulate extracting products 'a', 'b',and 'c' from its crude oil? Assume the company wants product 'a' on the west coast, 'b' in the middle of US, and 'c' on the east coast. Assume the company has refineries 'x' on west coast, 'y' in middle US, and 'z' on east coast. How does one model such a company's oil production so as to produce/refine the 'right' amounts of each product at each refinery site in order to meet the company's goal of maximizing profits?
Partial Differtential Equations (PDEs) will be used to model the crude oil distillation for each distillation unit at each site; i.e. many PDEs must be solved at once. Are there computers large enough to handle such problems today? Are there plans for some super computer that will be able to handle many (1,000s) PDEs at once?
With maintenance of distillation units being continual, e.g. fix one, stop another, this will be a constant problem when trying to simulate the next day's crude oil work load. For example, assume a company has 600 distillation units overall. That means a computer program would be required to solve 600 PDEs ASAP; i.e. 10 hours of PDEs. My past experience with modeling in FortranCalculus™ language/compiler, I was taught that a modeling requiring 'Tmod' time to execute the model, would require around 2'Tmod' time for the optimal solution. That would then get us into the 20 hr. time range for 600 PDEs. Too long! Need faster computer and solvers to get into reasonable solution times. Ideas how this could be done today?
Oil Refinery Production: An oil company may have many different distillation processes going on at each of its refineries. Each time new rude oil arrives at a refinery, or one distillation units goes in or out of services, a new tweaking of all distillation parameters must be done in order to maximize the company’s profit … from our textbooks.
This is an example problem written up in our textbook to get engineers thinking outside their box. It shows how one could 'tweak' which products to produce each time it is run. The code shown is for the FortranCalculus™ compiler. FC will tweak many parameters at once. Thus, it could be used to control the amounts of the various products being produced. Right now it would help Oil Companies from producing so much gasoline (where supply is high and demand is low) to other products such as black gold.
Findings: In the 1960s or 70s, the Chevron refinery at Richmond, CA implemented a computerized monitoring system at each of their control rooms. It was found that the avera
ge employee started their eight hour shift by tweaking their controls to setting that they new were safe. For the rest of their shift they read books or did other things of self interest. Then the computer monitor was turned on along with the plant manager telling these controllers that they could earn gold or silver or red stars as rewards for doing a good job of improving oil production. The computer monitors would ‘watch’ their tweaking. If they went into an unsafe zone for any control, it would stop them. After a few weeks most controllers were tweaking their controls to maximize some oil production and thus were receiving some gold/silver/red stars.
FortranCalculus Compiler for Calculus-level Problem-Solving; modeling, simulating, & optimizing. Learn to 'tweak' your parameters ... try it, you'll like it! Also, read more on this Oil Refinery example in our “Company Goal: Increase Productivity?” textbook (ex. 8-2).
Today, most individuals are working on getting their math model to provide an accurate description of one component of a project. We suggest moving past that and on to considering all components of a project or site or company. For example, those working at an oil refinery, they may be modeling one distillation unit. Why not consider modeling the whole refinery? This is now possible with a Calculus-level compiler. Finding the right objective may be an issue for total project simulation. This requires input from engineers to president of company.