Optimal Method of Distributing Edible Oil On Tanks

Well, no need to be rude, I am still student and I am trainee.

I will do a flow chart to know what is the solution. I am asking about a way for making this process automated for easier decision making for other people in the future. Management are suggesting nonlinear programming - operation research and simulation due to my engineering background.

so I will ask again, what is the best way to make it automated, after I draw the flowchart of course.

I would have green lights and red lights on the tanks controlled by centrally located computer, the lights would have numbers on them designating type of use....When a truck approaches the driver gives type of oil code, receiving or shipping, and time stamp...The driver then receives map with blinking light on proper tank and tank light of proper tank illuminates blinking....when the fill initiates the light goes steady on, and on completion goes out....completion of fill and quantity is sent to computer which updates inventory....green for receiving, red for dispersal...

The obvious solution comes to mind would be Mojo swoptops using the idea that possibly developed the shipping container. Each delivery of oil comes from the harbour offloading the tank; you no longer have to worry about the more complex distribution of the oil itself with the simpler tank distribution. Of course this is merely a wild idea, I think one would need an intimate knowledge of your operations to use a systematic approach to overcome the constraints, you seem to be familiar with the Theory Of Constraints (TOC).

It sounds like a typical logic puzzle. You might make a grid with tanks vs oil types and eliminate the incompatible grid squares.

http://logic-puzzles.org/how-to-solve-a-logic-puzzle.php

you might find it simple to assign each tank a particular score based on your current operation, such as receiving a truck load of raw oil. The tank with the highest score would win the shipment, then the score is updated and recalculated for all of the tank population. For instance, you would want to make sure the tanks in question suitable for raw oil have enough room for the entire tanker load, and rank them based on distance from the refinery. Not sure how shipment matters to the tank, unless you are only allowed to ship full tanks, but that could be added in. Zero out the scores for the incompatible tanks.

A maintenance screen would allow you to manually adjust the parameters based on leaks, cleanings, and other factors that allow you to reconfigure the basic tank compatibility, factors that cannot be updated by automatic sensors or calculations, or by the previous operation.

Rule based software such as that could be implemented with standard logic software, so your technical staff can understand it and troubleshoot it. If the rules are plain and simple, your operations people know how to modify it and tweak it. It seems that an elegant heuristic solution would perhaps not provide the value desired, if you need specialized skills to configure and maintain it...

It seems to me that the company you work for is asking more than your experience warrants. If you do solve the problem, you should demand a big raise and an office with a window, telephone and secretary; maybe even a company car.

Suppose then you have 28 tanks of which about half are for raw oil, and about half are for finished oil. You may not commingle raw and unfinished oil, and each tank has its own maximum holding capacity (a constant of the system), and each tank has it own present capacity (reported to the control room by instrumentation, hopefully redundant).

1. Get the blueprints for the plant layout, along with P&ID. Which tanks are interconnected (if any). Decide where to draw the boundaries for each product, and decide permanently how to isolate incompatible materials (cannot be next to each other without causing X to happen).

2. Firmly establish the pattern of load receiving/load dispatch so that no truck that is delivering product may approach tanks in the finished product area, and no truck that is taking on product shipment may approach the raw product holding area.

3. Connections to trucks may not be made by the drivers themselves. This is usually where chemical/product rules are broken. Each shipment incoming, and each load outgoing must have a tracking number for the shipment, does it not? Embedded in that tracking number (or maybe another field in the DCS) should be code for the type of material and its targeted quantity.

4. You can use other automated data control systems (DCS) used in warehouses for examples of how this plays out.

5. If tracking number code does not match the color or numeric coding on a tank connection, then it is forbidden to connect.

6. Metering control incoming controls inventory in the raw tanks, such that no tank should be allowed to overflow. In other words, if tank level leaves insufficient room for a T/L of oil, then the connection if forbidden. Some facilities may have rules that allow for "splitting the load" between incoming tanks, with pertinent cautions.

7. Load out inventory control is typically by weight on the truck, but some places still insist on doing this by level change on tank (not recommended).

8. Any plant's inventory control is only as good as the instrumentation within the plant, and the system that takes and applies the information.