Statistical
Pkuhocis
for
tne
J.nalysismid ?rediction
G. 2.McLaughlin
Section I
-t
t
Introduction:
In this paper we are concerned with estimating the
end results of patterns @f exploration activity in a region
having a sufficiently long history of discoveries of oil and
gas, so that it can be confidently asserted that the peak of
the discovery process has occurred, although the process of
discovery and appreciation of discoveries is far from complete,
Forecasts based
-2,
In this present study, we are analyzing patterns
of activity which are related to certain specific sedimentary
deposits, and it is shown that the learning process connected
with these sedimentary deposits is drawing to a close.
From
How-
be an in~lovativemove,
There exists
.O
By
appropriate trans-
Alternatively, it
may
are related to cumulative learning and we estimate the parameters involved from the discovery process itself using rlorllinear least squares techniques.
A further feature of the forecasts is an associated
set of confidence bands which are set using a mild non-parametric technique due to Gauss.
~Jatters
tical formulation.
Further discus-
\
,.
Since
by sediments to a depth in excess of 1000 feet but did not include the mountain ranges.
between the base of the glacial drift and the top of the Precambrian granite over the available acreage of sediments underlying the plains area of Alberta has been calculated as,
approximately, 225,000 cubic miles.
by assigning a uniform
.,*
-5Western Alberta) an additional 45,000 cubic miles of effectivebasin sediments may be calculated to exist in Alberta.(2)
In this domain have been drilled (as of December 31,
1970) 35$265 (3) exploratory outpost and development wells.
Of this figure, the total number of exploratory wells drilled
as of December 31, 1971 were 12,758, (4)
1.4
Discussion:
The subject with which we are dealing may be cha~acterized as trend analysis and the object is that of forecasting
the
process involved.
of time series analysis, and the second is based upon cumulative drilling.
Although,
(1)
*.
-7-
.~=-
L(t)*#(t)
dt
(z)
(3)
dt
is
a measure
knowledge.
We set
.
da(t)
1
dt
a(t)
k3>
(4)
(5)
** 8
-8have
2
~j~e
Y(t)= k1c2e3
-k ~
or
-k3t
dt
(6)
~e-k3t,
r
?3
= lc2e
(7)
(a)
(b)
(c)
1C2 = K
-k2
= lnA
c3k3
8(a)
-k3=
8(c)
lnB
8(b)
(9)
Note
Amount D~scover12d
in Time Period
61
&(l-i)
Is(l-i)l
6 (i-i)2
..*
b**
.,*
t
rs(l-l)t
d(l-l)t-l*i
(c)
Time Period
Amount Remaining
The constant A
may
$
Yi(tj:ls$ = 1,2,..., T
(lo)
I=ti
For subsequent ease of notaticm we now drop the subscript j
and write t = t.
$
***
10
(11)
as
lni?(t)= lnK + BtlnA + E(t)
(12)
n
Z bigi(t) + c(t)
i=l
(13)
bl
b2
(14)
in;(t)
= lnK
= lnA
gl(t)
(16)
(17)
p+(t) = Bt
In general, let
G =
(18)
e+)
II
132(t)
.
(19)
I@)
11
-11and
:1
bl
b2
.
.
bon
(20)
(12),
=
E(; -xb)(;
~ - b)t
~
(21)
UG2X
*
and that
2
%G(t)
where
;Ez
(22)
GZG
.-..,
T
;2
[(x(t) - ;P(t)]
t=tl
ql- n
z
(23)
and where
x(t) = lnY(t)
(24)
(25)
(26)
The value h is derived from Gauss inequality,
Pr{lt-
Ml}>h~~~h
4 *2
(27)
.* 12
- 12r= ~z + (p - m)2
wnere
(28)
(29)
where the parameters K, A ;md B have been estimated by nonlinear least squares techniques.
1.8
13
A suitable
This cat-
14
\
.,
-14reserves.
15
-l!j -
We interpret these
Year
Actual
1970
56549
56411
1971
58161
58351
Forecast
1972
60165
1973
61856
Acknowledgement:
Acknowledgement is made
Dr. M. Sablatash and with Ross C. R
the computer programs and the reali
.
..
REFERENCES
1.
2.
3.
Alberta Oil
4.
5.
6.
7*
8.
References (cont~d.)
9.
11
120
500
400
200
100
0,
VOL. 40, W
648-649,
10
TIME TO LEARN
15
ORIGINAL LEARNING
19S0
20
2
FIGURE
EFFECTIVENESS OF EXPLORATORYDRILLING
1955-58
REFERENCES
ALBERTA
DEPARTMENT
OF MINES
GOVERNMENT
OF THE
AND
1947-1970,
MINERALS,
PROVINCE
OF ALBERTA,
1971.
RESERVES
PROVINCE
DECEMBER
GAS
OF ALBERTA,
BOARD, CALGARY,
31,1970.
ACTUAL
FITTED
DATA
DATA
-%x
1967-70
TIME
cO!!SZZvA~ION
9CAnD(1968;
Yez? Of Estlnate
1952 1953
1951
488
204
195$
818
687
1955
1956
828
1957!
1958
826
826
803
1! 59
767
19GG
1961
1962
1963
77s
817
817
807
1964
:56[
1955
1966
1957
e22
666
839
15<8
816
1$)52
987
1953
15721
161;
Z?.?l
1766
1934
2131
:.158
2257
2389
2S25
2630
2763
2871
3510
3059
3.9..L
1022
1993
2630
2346
237b
2870
:135
3hl?
3?28
3818
3853
3941
4253
3949
3900
3683
548
937
Ikob
1951
2416
2751
3396
3215
3457
3Q43
3469
3596
3611
3603
35C7
236
1064
a35
? 36
:03a
loab
;188
]2a5
1322
1399
1438
1440
lb76
155E
457
S25
ao9
850
1372
1498
1655
1728
2049
220$
2239
24sa
2S27
2a93
1954
1955
1956
1363
1957
Ig~B
227b
:726
3263
3260
338a
299Q
2910
2791
2943
d
29538
1397
1901
2258
236.5
2552
255a
2512
2466
24a7
2501
2b95
71s
1S62
1773
17s9
la9c
2S72
35s5
3574
3832
3*5?
690
625
-aas
897
1087
1333
1585
163a
161S
:57
850
958
2000
2116
2267
2S92
4102
304
1160
1703
1666
1962
21s4
2147
320
392
4s1
sal
629
BG7
273
aea
867
697
716
487
622
a73
1017
307
306
303
368
1411
1959
1s)52
1361
[
1962
i963
19(34
I
1956
1965
.
I
i96i
I
1368
Totals
*
All
2ok
1*75
32al
4970
figures
givenin 3CF.
S3S0
7861
9a08
1s9
136+5
16s141
19392,
2i615
23295
24566
27829
3G9C6
32510
3*lla
36942
Year
Xscovezy
~e~r
1951
1551
1952
226
1352
1953
lW!
1955
:956
1!357
1958
:9~~
1963
19f~
778
2181
2280
2412
2S48
2623
2786
2789
3134
3307
3698
3688
3833
2076
154i
2912
3489
3373
3E15
1974
756
901
951
1001
1160
486
554
851
889
1418
1303
2214
26S8
1419
1925
860
848
831
9s9
1584
1623
1733
1778
1958
2154
1022
1993
2630
2319
2332
548
942
1454
1956
829
850
22Y
i96~
837
040
1955
1960
837
709
1954
1959
793
500
1553
1957
1958
789.
1959
6J19
1965
abti
1966
1967
1968
888
861
838
819
2896
3047
3146
3091
3141
3924
4237
3920
3879
3862
3077
3601
3642
37b9
3765
3757
3741
3606
125S
1292
13s4
1378
1462
1498
1580
1606
1520
168S
17.s8
2069
2223
2264
2479
2548
2550
3173
3270
3398
2904
2820
2701
28S3
2863
2802
271a
2201
2627
2628
2624
2564
2s18
2530
2S52
2546
2670
1496
177s
164S
1782
2458
3658
3477
373s
3761
196C
1961
:363
3782
i
691
818
878
890
1080
1326
1s78
1631
16:2
-1
1632
580
853
981
2023
2X49
2290
261S
4125
-t-
4800
304
1160
170s
1666
1962
2154
21&
21:1
323
392
4s1
581
828
806
=1- I
8c2
273
808
068
698
717
717
463
798
993
1962
1~~~
lC)~a
-.
:555
1993
13.12
1356
1.
:5.57
308
307
304
30s
35a
:402
1388
I
I
Lgca
282
1
:969
* AL
SS9
~~$als
226
figures
1499
given
331s
ECF.
50041
63811
78s9
9827
282.
12700
16288
!20009
21878
23?.34
24S97
2786?
31147
32599
343s4
3als7
38718
.-,. =-s
ALBERT.:I~;I.1IAL
E.STAbLLL1.
L FA3XkXA!E.i.
_.
GAS ?ZSZ?P3:
241
1?52
4
19?0I
1969
840
S41
836
819
782
786
83C
826
020
76a
835
879
e43
790
eo7
761
998
1s83
1629
1740
1810
199*
2174
2200
7264
2*I4
2S77
2670
285S
2913
2944
3062
3107
3148
3132
1024
2002
2230
2309
2312
2830
3078
j409
3666
3783
3801
391s
.215
3936
3882
3865
3984
4003
532
926
1838
2070
2S73
2320
3472
3339
3596
3582
3632
3739
3751
3741
3720
3*84
3124
97JI
826
991
981
1053
1240
1335
1372
1434
1458
15s2
1498
1574
1600
164S
63o
847
aas
1414
1529
1686
1760
2071
23o4
2263
2445
2504
Zkso
2460
1315
2145
2s83
316;
3188
3304
2832
2773
26S0
2762
2788
2727
2736
2737
1427
1933
2289
2297
2616
2s0s
2368
2459
2475
2547
2S6S
2687
2670
68b
1s17
1722
1819
1814
2bS7
3542
3s.11
3769
3786
3827
3981
781
908
958
960
1364
1410
1661
1703
1725
1695
1663
S81
800
1005
205s
2170
2315
2646
4)2S
4772 , 4925
314
1178
1721
1684
1983
2172
2165
2169
2174
32o
392
4s1
579
826
806
807
837
276
929
910
740
760
830
841
417
76o
749
07s
1194
1197
289
323
29S
386
141s
1188
1957
1958
1959
196o
1961
lgcz
1963
1964
1965
1966
1557
1
1
..-..
1968
1569
1970
1967 1968
830
1956
Totals
1965 1s66
709
229
1960
1964
520
1955
1958 1959
i96j
1955
1956 1957
1961 1962
1953
1954
Year of EstiRate
Discovery
Yezr
1951
1951
(1970)
241
All figures
:sla
3316
4993
givenIn BCF.
S965
a260
265
. .
II
99a3
307 \
33s
I 2099 : 238a
zal
391
1138067
sal ! z401
233
260S3
t
20]4a
21714
23694
24719
27871
31X76
1
32570
34118
37077
39461
4oa9a
I
1968
FUFKXIONSFOR INITIAL
FURECAST
ESTABLISHED
ALBERTAMARKETABLE
GAS RESERVES
WITH80% CONFIDENCE
LIMITS
YEAR
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
LOWERLIMIT
(TcF)
ACXUAL
(TCF)
4728,9
5039.6
5483.
?
6090.1
;::;.;
5158.1
5497.0
5981.5
6642.9
7509.6
8603.8
9112.8
10564,0
;::;:;
16192:1
18427,8
20797.1
23267.9
25807.4
28383.5
309661
33527.8
36044.6
38496.0
40865.4
43139.4
45308.1
47364,8
49305,3
51127.7
PREDI(XED
(TCF)
10305
11750
13802
15721
17289
19007
21219
9939.9
11522.8
13348.0
15401.6
17661,7
20100 3
22684.7
25576
28302
32792
34636
m;
25379.8
2814s?
.7
3oo;~
,;
41697
45194
46983
48844
52053
UPPERLIMIT
(TCF)
5626.2
5995*9
6524.3
;;;;.;
9384;7
10842.0
12568.7
;;;;;;
19264:7
21924.7
24743.6
;;;:;.;
53780.3
55768.1
33769:6
36842.2
39890.0
42884.4
45801.1
48620,0
51325,5
53905.8
56352.8
58661.5
60829.7
36570:8
39316.
I
41990.0
44574.4
47054.8
49420.4
5i663.7
1975
58772.0
63832.5
69328.7
1980
63681.2
69200.9
751991
1985
66797.6
726I.o.2
78928.7
1990
68721.6
74715.6
81232.3
1995
69889.9
?5994.2
82631.6
2000
70592.5
76763.1
83473.2
COMPARISON OF PREDtCTIONS
OF lNITiAL ALBERTA MARKETABLE GAS RESERVES
Cu= 800/. UPPER CONFIDEMX BOUND-1968
CL= 8070 -R
c~~~E
FIGURE
DATA
90
85
11.4
)0.4
80
8.9
ma
75
r
2.3
70
65
I#
LEGEND
1970
1%9
1968---55
I
-rim
1975
1980
Im
1995
..
1969
IKREC.4ST FUNOTIONS
FOR INITIAL
ESTABLISHED
ALBERTAMARKEIABLE
GAS RESERVES
WITH 80%COIWIDENCE
LIMITS
LQWERLm
(~)
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
4776.8
5098.5
5551.0
6161.4
6954.1
7948.4
9157.4
10586.5
12233.2
14087.2
16131.5
18343.3
20695.7
23159.3
25703.4
28297.6
30912.7
33521.7
36100.3
38627.3
41084.9
43458.4
45736.3
47910.2
49974.1
51924.8
:975
ACTUAL
(TCF)
..
10330
10777
13841
15760
17328
19004
21241
24505
28355
32811
34935
36589
38o25
41739
45468
47118
49112
53434
5407X
PREDICTED
UPPERLIMIT
(TcF)
(TCF)
5210.3
5683.0
6065.7
6604.1
5561.1
6054.7
6720.5
@ ;
9988:3
11547.0
13343.1
15365.4
1759592
20007.6
22573.5
25260.7
28035.6
30865.2
33717.6
36563,3
39375 9
42132.2
44812.7
7330.3
8273.3
9456.2
10894.6
12594.8
14553.8
16759.6
19191.7
21823.1
24621.8
27552.7
305795
33665.8
36777.0
39880.9
:;;4;.:
47401.6
4g886
.2
52257.3
54508.5
56635.9
48878:9
51702,6
54412.7
56998.9
59454.4
61774.9
60209.7
65418.5
71078. I.
1980
65709.4
71437.2
77664.3
1985
69296,2
75364.3
81963.7
1990
71567.4
7785I.. 5
84687.5
1995
72979.8
79398.7
86382.0
2000
73848,8
80350.5
87424.8
COMPARISON OF PREDICTIONS
OF INITIAL ALBERTA MARKETABLE GAS RESERVES
Cu= 30 0/0UPPER CONFIDENCE BOUNDCL= so:/o L~R
FIGURE 4
969 DATA
DATA
90
35.3
es
Cu
80
304
W.9
.0
Oz-
75
Gf---0
0
~6.8
> 7
9
70
LEGEND
1970
[969.
196e----
197s
1980
1990
1995
2000
z
NATIONAL EKRGY
w.
unme7.
..
1970
IWECAST
FUNCTIONS
FOR INITIAL
ESTAJ3LIS~ALBEFTAMARKETABLE
GAS RESERVES
W12H 80% CONFIDENCE
LIMITS
lXNIER
Lm
(m)
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
ACTUAL
(TcF)
4795.6
5119.9
5578.1
;5:; . ;
8017:9
9250.2
10706.6
12383.4
14269.1
16345,1
18587.0
20966.5
23ti52.
ti
26014.1
28619.1
;;;;:,:
36412;7
38923.0
4135795
43702.2
45946.3
48082.1
50104.4
52010.4
10347
11799
13842
15747
16852
lg460
21418
25762
28316
32969
34749
36999
38163
41745
45501
47o85
48844
52207
54916
56549
PREDICXED
UPPER
LIMIT
(T@)
(m)
5201.4
5553*1
6050.1
6722.5
7597.4
8696.2
10032.8
11612.4
13431.1
15476.3
17728.0
20159.6
22740.5
25437.1
282~5.
z
31040.5
33881.2
36707.6
39393e5
42216.1
44856.6
47399.7
49833.7
52150.1
54343*5
56410.8
5641,4
6022,9
6561.9
7291.3
51410,0
54049.9
565624
58941.3
61183.5
824o,2
$)h32.O
10881.7
12594.9
14567.4
16785*7
1922?.9
21865.2
24664.4
27589.2
30602.2
33666.7
36747.8
3981.3.3
42834,9
45787,9
48651.7
1975
6o04j.8
64882.3
70106.1
1980
65321,1
70618.0
76344,4
1985
68722*5
74319.1
80371.4
1990
70854,1
76639.1
82896.5
1995
72167.1
78068.4
84452.2
2000
72967.5
78935.
(
85400.3
..
.
COMPARISON W PREDICWONS
OF INI17AL ALBERTA MARKETABLE GAS RESERVES
II
FIGURE
90 +
85
)3.5
10.4
80
8.9
%6
75 r
4.7
LL
LJ
+
1,
*E 70
Ii!
65
60
1970
t669
----
55
s+
.lm
197s
Mr!ONALfwM?YH,30MAwMm
1690
1995
W.
MOI056-P
,~
FIGURE 6
2800
Z(t) s t@tog
Adog
B~Bt*A6+
2400
2000
id
.
1600
~
N
I
1200
800
NOTE : GRADH DERIVED
TECHNIQUES
RESERVES
400
LIQUIDS
0 ..-
1945
VIA
NON-LINEAR
APPLIED
SULPHUR,
ALBERTA
ENERGY
CALGARY,
ALBERTA,
PROVINCE
RESOURCES
IN
GAS
OF ALBERTA,
CONSERVATION
DECEMBER
BOARD,
3t, 1971
19s5
1965
197S
1985
t
..
SQUARES
TO DATA CONTAINED
OF CRU~E
AND
LEAST
19%
Mathematical Formulation
Let y(t) be an ergodic stochastic process and let
Y(t) = Pr{y(t) ~y}
Bt
= A
aol
Y(t) = o
t
= AB
t~o
a.2
a.3
u(t) = ft (S(t)dt
.m
a.4
where
so that
U(t)=l,
t~o
= o,
t<(l
a.5
Hence, we may obtain the density function of y(t) by differentiating equation a.3 with respect to t.
That is,
dF(t)
BtlnA*lnB*U(t) + A*6(t)
dt = P(Y(t)) = ABt*
a.6
= 6(t)
a.7
-2-
We may integrate equation a.6 obtaining, for t>O,
f~ ~F(t)
~t
(A
~Bt,Bt
-m
Bt
A)
a.8
~~p(y(t))dt
lnA*lnBU(t)dt + ~~ A*6(t)dt
U(t) + A
U(t)
a.9
a.10
=A
It
13t
follows from
all
that
~~ p(y(t))dt
d
~a~
F(t)
ABm
= 1
a.12