You’ve built your AFT Fathom or AFT Arrow model, defined all pipes and junctions, specified your fluid properties and defined your pressure boundary conditions. But for some reason, your predicted flows are still not matching measured data. AFT Fathom or AFT Arrow is properly solving the fundamental equations, but it is likely that something is off in the inputted data. In many cases, this is the pipe’s resistance. Many factors can alter a pipe’s resistance, but the two primary reasons resistance may be incorrect is (1) corrosion of the pipe over time, which increases a pipe’s absolute roughness and (2) the buildup of residue on the walls of the pipeline, known as scaling, which decreases a pipe’s inner diameter. This decrease in diameter is very important and will increase a pipe’s resistance dramatically. Now you have a dilemma. You can’t cut open your pipe to see how much scaling there is or how much your pipe has corroded. What is the condition of your pipe? Using AFT Fathom GSC or AFT Arrow GSC, you can find out!

What is GSC? GSC stands for Goal Seek and Control and is an add-on module for both AFT Fathom and AFT Arrow that automates the search for input parameters to achieve desired output. In other words, it allows the user to define a set of model input variables that are changed until a set of output goals are reached.

The first step toward calibrating your model using GSC is building a completely defined model. In my AFT Fathom example model, I have an oil pipeline with two 50-mile-long gathering lines that feed into a single 50-mile-long delivery line.

It is known that only 32,000 barrels/day are flowing through line A, but the flowrate AFT Fathom is predicting is above 36,000 barrels/day. It appears that residue has built up during the pipe’s lifecycle, and the decreased diameter of the pipeline is restricting the flow through the line.

The flow in line B is predicted to flow above 29,000 barrels/day, but measured data is showing that flow is only 23,000 barrels/day. Here it would seem that scaling has also occurred.

b2ap3_thumbnail_Pipeline-Overview2.png

Using the “Goal Seek and Control Manager” window—available in the Analysis menu after GSC has been activated—the user can define their variables and goals. In this specific case, the variable is “ID Reduction (Scaling).” This will vary the inner diameter of both Line A and Line B until the output matches the specified goals. 

b2ap3_thumbnail_GSC-Variables.png

Here it is important to note that each pipe’s scaling is listed as a variable, but the scaling of the individual line is linked to the longest pipe in each line respectively. These pipes are P4 for Line A and P8 for Line B. This essentially creates two variables: the scaling in Line A and the scaling in Line B. This is mathematically important because we will have two goals, and the number of goals is required to match the number of variables.

b2ap3_thumbnail_GSC-Goals.png

Here the two goals are the measured flow rates in Line A and Line B. Any one pipe in Line A could be the Object ID for the 32,000 barrels/day goal, and similarly, any pipe in Line B could be the Object ID for the 23,000 barrels/day goal. Pipes 4 and 9 were chosen for simplicity, as they are the pipes that connect to the tee where the two lines combine. After the variables and goals are specified, the Workspace will appear as shown below:

b2ap3_thumbnail_GSC-overview_20160216-004723_1.png

Each object that is used as a variable will have a red V next to the object’s ID, and if the object is a goal, it will have a red G.Now that GSC is completely specified, the model can be run again. It will take longer to run than a normal steady-state run, but that is because AFT Fathom is testing the possible combinations of input variables that will achieve the specified goal. Once AFT Fathom converges, you can view the GSC Variables Summary and GSC Goals Summary in the Output Primary Window.

b2ap3_thumbnail_GSC-Output-variables.png

The GSC Variables Summary displays the final amount of ID reduction that occurred in the two lines. It can be seen that the inner diameter of Line A has decreased 15.23% from the original pipe. Line B had even more scaling with a 21.77% reduction.

It would seem a substantial amount of scaling has occurred. Possibly too much. Could it be that we are forgetting something important? What about the scaling in the 50-mile-long delivery line? It is logical that if scaling has occurred in the gathering lines, then it is also likely in the delivery line.

If we want to add scaling in the delivery line as a variable, we need a new goal to keep a 1:1 ratio of variable to goals. The first thought might be that we could use flow again. We know the flow in the line; why don’t we use that? It is possible to define a model this way, but the results would not be unique. This is because the flow has already been defined in the delivery point by the other two goals. The flow already has to equal the sum of the two flows because of conservation of mass. A new goal is required. This means another measurement is required. If flow is already known, then pressure is a great measurement to take.

Downstream from the tee, there is a pressure gauge reading 965 psig. This measured pressure is an excellent goal to use, as it is a reading taken directly from the line we are trying to calibrate. The new variables and goals will appear as below: 

b2ap3_thumbnail_GSC-Variables-2.png

b2ap3_thumbnail_GSC-Goals-2.png

After rerunning the model with these variables and goals, it is easy to see that the scaling in the delivery pipe does in fact have a significant effect on the overall system’s scaling.

b2ap3_thumbnail_GSC-Output-variables-2.png

Adding scaling in the delivery line (P10-13) significantly reduces the scaling in the gathering lines. Line A’s scaling is reduced by more than 9%, and Line B’s scaling is reduced by more than 11%. This large scaling reduction in lines A and B at the same flowrates shows that scaling in the delivery line is much more impactful than scaling in the gathering lines. This is because the combined flowrate has a higher pressure drop given the same internal diameter. Knowing how much scaling is in each line gives the engineer a better idea of how to move forward and find the best solution.

What are the possible ramifications of this study? One possible result is financially justifying pigging the line. You have the math to say, “We could get more than 10,000 extra barrels per day by just pigging the line.” Perhaps the current flowrates are acceptable, but using this model, you could monitor how the flow degrades over time and better plan out when the line should be cleaned. Another possibility is after pigging the lines, one of the lines is still failing to deliver the predicted flow. This could be the result of a leak or a highly corroded pipe, and using GSC could help identify the problem area.

In summary, AFT Fathom GSC and AFT Arrow GSC are great tools to help augment the already powerful capabilities of the software. They continually improve the efficiency of the engineer using the software for increased sustainability and operational benefits.