Commercial Navigation Technical Working Group

munchsistersAI and Robotics

Oct 17, 2013 (3 years and 9 months ago)


Commercial Navigation Technical Working Group

This Technical Working Group (TWG) will improve our understanding of the relationships between
transportation costs and water levels and flows. For example, ships carry lighter loads in order to
compensate for low levels in connecting channels, requiring a
dditional trips and increasing costs.
Likewise, higher levels allow for deeper draft vessels carrying heavier loads. Flow variations at the Sault
Ste. Marie power plans also affect commercial shipping interests, with high flows during the day
benefiting na
vigation and lower flows at night and on weekends potentially delaying ship transits and
affecting cargo capacity. This TWG will examine such issues as “peaking and ponding” and work with
other TWGs to determine system
wide benefits and detriments and iden
tify changes to regulation plans
or criteria that might improve operations for navigation interests. A key focus will be climate change and
variability and how commercial navigation may need to adapt to respond to more extreme conditions
than experienced i
n the recent past.

I thought that the best way to validate or provide "stamp of approval" for the SVM was to see how well
it accurately reflected Great Lakes commercial navigation impacts based on changed water levels
resulting from any proposed Lake Su
perior regulation plan as it compared to the same runs shown in our
Great Lakes model (GL
SAND). I never expected that SVM and GL
SAND would have identical results as
measured by commercial navigation transportation costs because of inherent differences an
limitations we had between the two approaches but I did expect that if the transportation cost
relationships that I provided to Bill were incorporated correctly, that the results would be comparable
and that the SVM could be considered as an accurate ref
lection of changes in transportation costs
(economic impact) between proposed regulation plans and Plan 77A.

We have compared four 109 year simulations of the GLSAND and SVM (77A, 129, Nat64D, and 55M49,
historic net basin supplies) and found that while
the estimates of total costs produced by the two
models were different ($2.6 billion in GLSAND and $2.8 billion in the SVM) the estimates of net benefits
for commercial navigation, which is the difference in costs between plans were very similar. Both
els showed that the benefits for Nat64D and 129 were so small (both models estimated about
$50,000 per year for both plans, or about 0.002%) that they should be considered as indistinguishable
from Plan 77A. Both models showed that 55M49 would have a smal
l, but noticeable negative impact.
GLSAND estimated impacts as negative $1.5 million (
0.06%), the SVM estimated negative $2.2 million (
0.08%). Confidence in these numbers can be drawn not just from the similarity of the model results but
also the reaso
nableness of the outcomes. The water levels in plans Nat64D and Plan129 are very similar
to 77A, while 55M49 allows slightly lower Michigan
Huron water levels which impacts most ships that
travel on Lake Superior, and the negative results is consistent wi
th that.

These small differences between GLSAND and the SVM are likely due to necessary structural differences
between the models. For instance, GLSAND distributes costs from light loading differently from the SVM
and the SVM considers channel depths in
the St Marys River, whereas GLSAND considers just lake levels.
Actual future shipping costs will be different from either model. We believe that the SVM benefit
estimates will properly support a decision on a new regulation plan.

The SVM has a wide rang
e of statistics on navigation costs that should help the Board make a sound
decision. We believe the use of average annual navigation benefits is essential in decision making, but
the impacts of plans in the extreme should also be considered. Accordingly
, we hope the Board
considers the worst year’s navigation costs for each plan as well as the average.


Does the SVM emulate the GLSAND appropriately, given the decision framework?


Does the SVM display the right information for making the decision?


Is the Board using the information correctly? (You can offer advice to them)