60
Statistical Methods for Estimating Petroleum Resources
The degree of uncertainty in the prediction intervals is con3.
trolled by four factors: (a) the uncertainty inherited from the
superpopulation, (b) the ratio of the number of discoveries to
total number of pools, (c) the difference in reserves between
the two nearest pools, and (d) the individual pool-size distributions computed from specifi ed discovery records, which tend
to be less skewed and more concentrated around the medians
than those computed without specifi ed conditions.
Estimating pool sizes constrained to a discovery record serves not
only to estimate remaining resources, but also to reduce the uncertainty
inherited from the superpopulation. The current assessment is different from that conducted in 1988 (Podruski et al., 1988), as follows:
The conditional play potential is 26 × 10
1.
6 m
3 for this assessment, compared with the expected potential of 60 × 10
6 m
3 in
the 1988 assessment.
The largest remaining pool size could be as large as 3 × 10
2.
6 m
3
compared with the 1988 assessment of 7 to 15 × 10
6 m
3
.
The total number of pools is 400 in this assessment, compared
3.
with 60 in the 1988 assessment.
The reasons for these differences include the following: First, in the
1988 assessment, the Swan Hills A & B pool was considered a single
pool with reserves of 303 × 10
6 m
3
. For the current assessment, this
pool was divided into two pools (221 × 10
6 m
3 and 69 × 10
6 m
3
). And
second, the minimum pool size adopted by this assessment is about 10
times smaller than that of the 1988 assessment. Therefore, the N value
increases accordingly. The current prediction is not distorted because
of including noncommercial pools.