The distribution of rates is used to generate the range of production history lines corresponding with P90, P50, P10 cases. (For example, for each period, the P90 point is determined as the value for which 90% of the data points are higher—or for the more common description, the value for which 10% of the values are smaller.) The data points in this case are the rates for each well at a specific period. If a group of groups is selected, the aggregate rates for each group are used.
To determine this value, all data points for a period must be sorted in ascending order. N is the number of data points in the period. P is the percentile. (In the P90 example P is 10.) The position of the percentile value is n.
The value of n is split into its integer (k) and decimal (d) components such that n = k + d
The percentile value (vp) is calculated by:
The production volume for each period is calculated in the same way as the rate.
The cumulative production track is calculated by summing the calculated volumes for each period.
For example:
We have the following dataset:
| 1.18 | 1.28 | 0.95 | 2.49 | 1.35 | 1.82 | 1.46 | 3.96 | 1.33 | 1.40 |
| 3.72 | 3.11 | 2.32 | 2.20 | 1.75 | 4.57 | 2.22 | 4.12 | 2.79 | 3.95 |
The values are set to ascending order:
| 0.95 | 1.18 | 1.28 | 1.33 | 1.35 | 1.40 | 1.46 | 1.75 | 1.82 | 2.20 |
| 2.22 | 2.32 | 2.49 | 2.79 | 3.11 | 3.72 | 3.95 | 3.96 | 4.12 | 4.57 |
The percentile to be found is 10 (P90). N=20.
Next, n is determined:
From n = 2.9, k = 2 and d = 0.9.
Then we find v2 is 1.18 and v3 is 1.28.
Finally vk + d(vk+1 – vk) = 1.18 + 0.9(1.28 – 1.18) = 1.27.
The P90 rate for this period is 1.27.