Environmental Engineering Reference
In-Depth Information
(a)
6
Analog field: 20 good wells out of 100
5
New field: Eight good wells out of 10
4
3
P (New = Analog | Data from new) = 0.50
2
1
0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
θ: Frequency of good wells in the new field
(b)
9
8
7
6
5
4
3
2
1
0
Analog field: 20 good wells out of 100
New field: Eight good wells out of 10
P (New = Analog | Data from new) = 0.02
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
θ: Frequency of good wells in the new field
Figure 13.45 Example updated probability distributions for frequency of good wells in developing an uncon-
ventional gas reservoir.
of good wells based on a large sample of data is 0.20 while the frequency of good wells in the
new field needs to be >0.83 to be profitable. However, it is possible that information from
the analog field is not directly relevant to the performance of the new field since the develop-
ment technology is advancing rapidly and the reservoir conditions may not be identical (i.e.,
the new field could be a “Black Swan” in the words of Taleb 2007).
The principle of Decision Entropy Theory to maximize the entropy of what might or
might not be learned about the decision outcomes suggests starting with a 50-50 prob-
ability that the data from the analog field are relevant to the performance of the new field
(Habibi et al. 2014). Figure 13.45 shows the updated distribution for the frequency of good
wells in the new field based on the analog field and test wells drilled in the new field. If there
are no test wells in the new field, the prior probability distribution reflects the data from the
analog field and the prior probability distribution for the frequency of good wells in the new
field ( Figure 13.45 ). As data from test wells in the new field become available indicating a
much higher frequency of good wells than that in the analog field, the probability that the
analog field is relevant to the new field decreases and the updated probability distribution
for the frequency of good wells in the new field more strongly reflects the data from the new
field than the analog field ( Figure 13.45 ).
The value of information versus the number of test wells in the new field is shown in Figure
13.46 . For small numbers of test wells, there is no value to the information because there will
not be enough information to justify developing the play even if all of the wells are good.
The value of information from test wells in the new field assuming that the analog field is the
same as the new field is also shown for comparison in Figure 13.46 ; there is no value to this
information even with 50 test wells. A simple explanation for this result is to consider the
 
Search WWH ::




Custom Search