Table 24. Final Monte-Carlo time and cost estimates for hypothetical double-capacity cargo
ships.
Elapsed
Std.
Mean
Std.
Escort
Std.
Total cost
Std.
time
dev.
speed
dev.
time
dev.
($)
dev.
April
Noril'sk
561.15
37.29
5.57
0.30
514.55
40.26
1,307,535
65,969
Lunni
565.80
37.96
5.55
0.30
518.70
41.90
758,162
33,257
Strekalovsky
565.32
40.04
5.55
0.30
517.57
40.06
690,886
28,372
June
Noril'sk
510.34
29.21
6.67
0.38
175.89
54.01
1,219,659
52,605
Lunni
513.66
28.21
6.61
0.35
180.23
52.29
713,026
25,114
Strekalovsky
516.46
30.73
6.59
0.38
176.42
55.63
656,487
21,956
August
Noril'sk
309.74
31.59
10.19
1.06
75.71
39.88
822,884
56,361
Lunni
339.88
27.86
9.25
0.76
76.49
40.00
503,372
24,579
Strekalovsky
339.60
26.66
9.27
0.71
71.93
37.62
466,855
19264
October
Noril'sk
471.28
38.36
6.77
0.52
125.18
43.30
1,106,968
67,430
Lunni
495.11
34.93
6.45
0.41
124.30
41.73
638,325
3,0616
Strekalovsky
499.08
37.68
6.41
0.45
119.78
43.90
578,344
26,284
over the entire voyage from port of origin to port
computer model was a method for organizing and
of destination, that is, for example, from Yokohama
quantifying the extensive data that were assembled
to Rotterdam. Additional open-water distance off-
during the reconnaissance study.
sets the higher costs associated with the NSR por-
Interest in the NSR is currently high and there
tion of the voyage and reduces the overall average
have, no doubt, been proprietary efforts to model
transit cost.
its utility. To our knowledge, however, there was
Extending these model results to estimate the
no such software available in the public domain
total origin-to-destination shipping costs was the
that could be used for Corps reconnaissance pur-
ultimate purpose of the NSR reconnaissance study
poses. Some recent empirical data on NSR
and the role of the Alaska District. We invite the
trafficability have been published by the Interna-
reader to review the District's full reconnaissance
tional Northern Sea Route Programme (INSROP)
report (USAED 1995) to see how this modeling
and Russia's NSR Administration in an effort to
effort was incorporated into the overall result. Al-
foster greater international interest. These data
ternatively, a summary of the reconnaissance study
have primarily been in terms of average ship
is available as Smith (1995). Here, we shall simply
speeds for various routes, months of the year, and
state that these results, when spread over full ori-
broad categories of ice conditions. The problem
gin-to-destination transits between northern Eu-
with making projections based on these data is
rope and the Far East, did promise more economi-
that few foreign ships have made the voyage and
cal per-ton transportation rates than can be realized
complete information on their experiences is not
with today's ships.
readily obtainable for analysis. To overcome the
scarcity of data, we employed a MC method to
select the environmental conditions encountered
on a voyage and predicted transit time based on
CONCLUSIONS
expected vessel speeds under those conditions.
As part of an 18-month reconnaissance study,
Russian researchers have collected weather, ice,
the U.S. Army Corps of Engineers constructed a
and oceanographic data in the Arctic Basin for
Monte Carlo-based FORTRAN model to calculate
many decades and they have developed probabil-
transit time and costs for cargo shipments between
ity-of-occurrence relationships for a multitude of
Murmansk and the Bering Strait using the North-
environmental parameters that affect polar navi-
ern Sea Route. The model enabled the Corps to
gation. These extensive data, published in atlases,
compare the costs of shipping via the NSR with
monographs, reference books, and articles were
those of alternative routes, which then allowed a
the cornerstone of our model. The large amount
prediction of future commodity movements. The
and form of the Russian data on the route's envi-
46
Previous Page
