About the Production Analysis Plot

When you create a Production Analysis, Meridium APM plots the production output values. The output values themselves are plotted along the x-axis, and the Reliability Percentage values for each datapoint are plotted along the y-axis. The Reliability Percentage is calculated by first determining the median rank of each output value and then using the median rank in a second equation that converts it to the Reliability Percentage. The following simplified equation can be used to determine an approximated median rank for each output value:

Median Rank=1-(i-.3)/(N+.4)

...where:

i = The order of each datapoint with respect to other datapoints. To determine the i value of each datapoint, the output values are sorted from highest to lowest. The highest output value is assigned a value of i=1. The second-highest value is assigned i=2, and so on until all the output values have been numbered.
N = The number of output values. For example, if your Production Analysis has a year's worth of datapoints, then N=365. If your Production Analysis has only a month's worth of datapoints, then N=31 or N=30, depending on what month you are using.

After the median rank has been calculated, the Reliability Percentage is calculated using the following equation and then plotted along the y-axis:

Reliability Percentage=100-(Median Rank*100)

The general shape of the plot can provide you with basic information about your production process:

Ideally, the datapoints will be plotted to form a vertical line at the highest possible X-value. This would mean that your process is producing the highest rate of output at all times, with no variation, and that you have no reliability or process problems. While this scenario is not realistic, visualizing the ideal plot compared to your actual plot can be useful in understanding how far your actual data deviates from ideal data.
If the datapoints are plotted in a straight, diagonal line that slopes downward from right to left with no sharp change in production rate, then any variation in output is due to process problems; you have no equipment/location reliability problems.
If the datapoints show a sharp change in production, then you have both process problems and equipment/location reliability problems. You will draw your Process Reliability where you notice this sharp change in production output.
Within the reliability loss region, you may notice distinct areas where the production output varies greatly from the rest of the datapoints in the region. These variations may be due to different types of reliability problems, such as system shutdowns, intentional cutbacks, and equipment/location failures. You can create breakpoints to divide the reliability loss region into separate areas where you notice these changes.