Professional Documents
Culture Documents
To search for other technical reports published by ERDC, visit the ERDC online library
at https://erdclibrary.on.worldcat.org/discovery.
ERDC/CHL MP-23-2
December 2023
Final Report
Abstract
The US Army Corps of Engineers (USACE) traditionally uses two metrics
to evaluate the maintenance of coastal navigation projects: tonnage at the
associated port (representing relative importance) and the controlling
depth in the channel (representing operating condition). These are
incorporated into a risk-based decision framework directing funds where
channel conditions have deteriorated and the disrupted tonnage potential
is the highest. However, these metrics fail to capture shipper demand for
the maintained depth service provided by the USACE through dredging.
Using automatic identification system (AIS) data, the USACE is pioneering
new metrics describing vessel demand for the channel depth, represented
by vessel encroachment volume (VEV). VEV describes the volume of the
hull intruding into a specified clearance margin above the bed and
captures how much vessels use the deepest portions of USACE-dredged
channels. This study compares the VEV among 13 ports over 4 years by
combining AIS, tidal elevations, channel surveys, and sailing draft. The
ports are ranked based on the services demanded by their user base to
inform the decision framework driving dredge funding allocations.
Integrating demand for-depth metrics into the Harbor Maintenance Fee
assessment and/or Trust Fund disbursements could alleviate the
constitutionality concerns and several criticisms levied against Harbor
Maintenance funding.
DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes.
Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products.
All product names and trademarks cited are the property of their respective owners. The findings of this report are not to
be construed as an official Department of the Army position unless so designated by other authorized documents.
DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR.
ERDC/CHL MP-23-2 iii
Preface
This work was conducted for the USACE-Jacksonville District under the
Scope of Work for Port Connectivity and Under-keel Clearance for South
Atlantic Division Ports. Processing and analysis were funded by the
USACE Coastal Inlets Research Program.
offset—Fig. 3). Then, the metric reported as the VEV for a given Research Area
vessel position report is the volume of the vessel hull that intrudes
into (encroaches) on the fictitious bed domain region—the space This study encompasses the eight high-tonnage coastal ports and
between the true channel bed and the bed domain offset. As with five additional coastal ports within the USACE-SAD. Table 1 dis-
ship domain violations, the VEV reported is necessarily dependent plays the average tonnage at each port between January 1, 2015,
on the selection of the bed domain offset—effectively the minimum and January 1, 2019 (USACE-WCSC 2021), sorted by average ton-
allowable margin of underkeel clearance. nage rank over the study duration from the largest to the smallest.
Mariners and waterway managers approach the available depth These 13 ports accounted for $109.1 million of average annual
constraint from opposite perspectives. Mariners maximize profit by dredge expenditure for the studied period. Fig. 1 shows the loca-
maximizing vessel draft but risk keel strike by loading vessels too tions of the ports.
aggressively. Waterway managers maximize global system perfor-
mance by minimizing local maintenance costs by providing mini-
mally adequate depth. The present study suggests that two bed Methods
domain offsets can be used: one to demarcate vessels with an ele-
vated potential risk of keel strikes due to the extreme proximity of Calculation of the VEV for multiple ports is a large-scale data fu-
the bed and one to identify vessels at low risk of disruption that are sion exercise from multiple open-source data sets. Marine Cadastre
still using the majority of the available depth in the channel—the AIS data (BOEM and NOAA 2019) are the source of the time-
Goldilocks region. The bed domain offset used to quantify the stamped vessel location, horizontal vessel dimensions, and heading
VEV in the elevated risk region is informed by the PIANC recom- information, with a temporal resolution of 1 min. The available
mendation for the maneuverability margin in full-depth channels channel depth during the vessel position report is computed by
that are not impacted by significant wave action—“the greater of summing the eHydro survey depth data (USACE 2020B) with
61 cm or 5% of sailing vessel draft” (PIANC 2014). To demarcate the tidal elevation from the nearest NOAA tide gauge at the time
the elevated risk VEV to maintain consistency across vessels/ports of the vessel transit. The temporal frequency of USACE channel
and to ensure a uniform change in the resulting sensitivity to bed surveys varies from daily to annually across projects due to differ-
domain offset selection, 61 cm was ultimately selected. For the ences in bathymetric survey practices (Scully and Young 2021) that
ports in this study, the selection of 61 cm versus 5% of vessel reflect differences in the underlying morphodynamic behavior of
draft has little effect on the results. The authorized depth of the the bed for individual USACE navigation projects. The vessel
deepest navigation channel in this study (Mobile, Alabama en- draft at the time of the vessel position report is obtained from for-
trance channel) was 13.72 m (USACE 2020b). The difference be- eign vessel entrance and clearance data published by IWR
tween 61 cm of clearance and 5% of vessel draft for a vessel (USACE-IWR 2018). The process by which these data are inte-
drafting 13.72 m is 8 cm, well below the USACE surveyed depth grated into VEV results is depicted in Fig. 2. More details on the
uncertainty (Byrnes et al. 2002). A bed domain offset of 1.52 m underlying data sources, associated processing steps, and filtering
was used as the upper bound of the Goldilocks region. This selec- criteria are found in the study by Scully and Young (2021). The
tion is not supported by formal guidance but was informed by the novel methodology presented in this paper is the additional compu-
findings in the study by Scully and Young (2021), who demon- tation of the VEV (Fig. 3) and comparison among ports.
strated that most vessels transit with clearances well above
PIANC recommendations. In that study, 75% of the vessels draft-
ing over 13.72 m in the Southwest Pass, Louisiana, transited with Calculation of the Vessel Encroachment Volume
1.52 m of clearance or more. VEV quantifies how much vessels
are utilizing the full depth of the channel and can be easily aggre- The VEV was calculated for every vessel position report identified
gated by spatial extent/across time to compare demand for available within the navigation channels of the 13 ports. To compute the
clearance across ports. Using two distinct bed offsets allows this VEV for one vessel position report, the “vessel shell,” defining
demand to be segregated by risk. the horizontal spatial extents of the reporting vessel, was
Fig. 1. (Color) Map of the 13 SAD ports in this study. (Map from Esri, HERE, Garmon, FAO, NOAA, USGS, EPA.)
constructed from the vessel’s horizontal dimensions and heading in MLLW (positive higher and negative lower); thus, the depth avail-
AIS. The eHydro survey points that fell within this vessel shell able to the vessel at each survey point beneath the vessel shell is the
polygon were determined by spatial intersection. The time stamp sum of the depth and the tidal elevation. For each vessel position
of the vessel position report was used to determine the most tempo- report, the available depth at the ith survey point intersecting the
rally proximate survey elevation for each of the survey points iden- vessel shell was defined as the sum of the surveyed depth zi and
tified within the vessel shell. This allowed the identification of the the time-interpolated tidal elevation ζi(t) at the nearest
surveyed bathymetry beneath the vessel at the time of the vessel gage (Fig. 3). The VEV for the ith survey point is calculated as
transit. The survey data are reported as positive down from mean follows:
lower low water (MLLW), and the NOAA tidal elevation data
are reported as the water surface elevation deviation from
VEVi = [d(t) − (zi + ζ i (t) − CBDO )] × A (1)
VEVpr can be aggregated over any desired temporal and spatial do-
main (in this case, to determine the average annual VEV for each
port). Note that the VEV does not represent the volume of sediment Fig. 4. (Color) Position reports for vessel transit outbound from
necessary to be removed from the channel to eliminate shoals in the Mobile, Alabama. The background color is the available channel depth.
path of vessels. Multiple vessels may encroach on one shoaled sur- Red-shaded quadrilateral icons indicate underkeel clearance. The dashed
vey cell, or the same vessel may encroach on a shoaled survey cell black line shows the channel cross section displayed in Fig. 5.
multiple times (e.g., during its inbound and outbound transits).
Rather, it describes how many vessels transiting a given waterway
are using the deepest portions of the available channel depth (i.e., and riverine inputs may have a more significant effect on the chan-
the volume of all vessel hulls transiting within the bed domain off- nel depth and are not included in the NOAA predictions. For ports
set from the bed). VEV results are directly related to the selection of with larger tidal ranges, the tidal elevation contribution is more sig-
CBDO; thus, VEVs reported in the results will be preceded by the nificant (e.g., Savannah, Georgia—Table 1), and there are ramifica-
associated CBDO value (61 cm VEV refers to the VEV for a tions related to the draft of vessels able to call at ports with similar
CBDO = 61 cm). The advantage of VEV over raw counts of en- maintained depth but different tidal ranges. The positions of the
croaching vessels is that it incorporates the effects of (1) the se- 61 cm and 1.52 m bed domain offsets are shown as dashed red
verity of the encroachment and (2) the size of the encroaching and dashed-dotted blue lines, respectively. The intrusion of the ves-
vessel. The advantage over quantifying the volume of sediment sel hull (solid green outline) into these 61 cm (red region) or 1.52 m
necessary to remove to eliminate shoaling is that it incorporates (blue + red region) bed domains represents the vessel encroach-
the number/size of vessels impacted by the shoaling. ment area for this cross section. Integrating this over the length
of the vessel results in 1.52 m VEVpr = 5,707.54 m3, the volume
of intrusion into the 1.52 m bed domain for this position report.
Results Vessel hulls are simplified by representing them as rectangular
prisms with depth equal to the vessel draft. The accuracy of this
An example of the raw results is shown in Fig. 4, displaying posi- representation depends on the individual block coefficient of
tion reports of a cargo vessel outbound from Mobile, Alabama each vessel (Rawson and Tupper 2001). The precision of the results
drafting 13.72 m in the northern portion of the entrance channel may be improved by modifying the hull shape based on vessel
of Mobile Bay. The background color is the available depth in class; however, this improvement must be weighed against the
the channel (channel depth plus tidal elevation), and the overlaid more intensive computations required to discretize the vessel
vessel icons (quadrilaterals) are shaded red based on underkeel draft over the hull.
clearance for the given vessel position report. White vessel icons Figs. 4 and 5 provide qualitative descriptions of the factors im-
have a clearance greater than 1.52 m. This vessel transit is the pacting each vessel in the channel. As with the underkeel clear-
sixth observed in this section of the channel for the week of ances in the study by Scully and Young (2021), the VEVs in
February 11, 2018. The figure text notes the aggregate 1.52 m these figures describe the available depth demand of individual ves-
VEV for all transits during this week until the time specified in sel transits in selected sections of single channels. Aggregating
the figure (the first six transits in the week of February 11, 2018). these values for comparison across ports makes them valuable as
These results can be viewed in cross section to visualize the management metrics. To that end, Table 2 displays the average an-
61 cm and 1.52 m VEVs for specific vessel position reports. The nual 61 cm VEV, the average number of 61 cm encroaching tran-
dashed black line in Fig. 4 shows the position of the channel sits (ETs), and the average 61 cm VEV normalized by the
cross section displayed in Fig. 5. average number of ETs for each port. It also displays each port’s
Fig. 5 highlights the contributions of the input data to the result- respective ranking by average 61 cm VEV and ET-normalized
ing VEVs for Mobile Bay. Gulf coast ports like Mobile experience 61 cm VEV, as well as the difference between those rankings
relatively small tidal ranges, limiting the impact tidal elevation has and the port’s ranking by average tonnage. The ports are sorted
on available depth. In such microtidal channels, atmospheric effects by tonnage, as in Table 1.
Table 2. Average 61 cm VEV (Mcy X T), average 61 cm ET, average 61 cm VEV per ET (m3), rank by average 61 cm VEV (difference in rank from rank by
tonnage), and rank by average 61 cm VEV per ET (difference in rank from rank by tonnage)
Avg. 61 cm VEV Avg. 61 cm ET Avg. 61 cm VEV per ET Rank–61 cm VEV Rank–61 cm VEV per ET
Port (Mm3 X T) (#) (m3) (Δ) (Δ)
Mobile, Alabama 4.47 125 35,760 2 (−1) 1 (0)
Savannah, Georgia 5.50 459 11,983 1 (+1) 4 (−2)
Tampa, Florida 1.09 59 18,474 3 (0) 2 (+1)
Pascagoula, Mississippi 0.01 6 1,372 10 (−6) 11 (−7)
Port Everglades, Florida 0 3 159 13 (−8) 13 (−8)
Charleston, South Carolina 0.35 31 16,436 5 (+1) 3 (+3)
Jacksonville, Florida 0.06 14 4,025 8 (−1) 9 (−2)
San Juan, Puerto Rico 0.33 28 10,317 6 (+2) 7 (+1)
Miami, Florida 0 3 334 12 (−3) 12 (−3)
Wilmington, North Carolina 0.32 26 11,470 7 (+3) 5 (+5)
Canaveral, Florida 0.01 3 11,449 9 (+2) 6 (+5)
Brunswick, Georgia 0.66 141 4,227 4 (+8) 8 (+4)
Palm Beach, Florida 0.01 2 1,481 11 (+2) 10 (+3)
VEV results can be considered through the lens of either total fewer ETs. Thus, the risk of the average 61 cm VEV per ET in
demand (average 61 cm VEV) or demand per transit that en- Mobile, Alabama is elevated, indicating higher demand placed
croaches on the bed domain (average 61 cm VEV per ET). These on the full depth on the part of a smaller subset of users. The
metrics give the USACE a description of the demand navigation 61 cm VEV also highlights ports with low demand for the existing
channel users place on the channel maintenance service provided channel depth. Pascagoula, Mississippi, and Port Everglades, Flo-
through USACE dredging at their respective projects, including rida, were consistently ranked among the lowest ports in SAD by
how that demand is distributed. For example, Table 2 shows that 61 cm VEV metrics, yet they are the fourth and fifth largest ports
large numbers of vessels demand the full channel depth at Savan- by tonnage.
nah, Georgia, which has the largest overall 61 cm VEV and the
largest number of average ETs. However, the severity of the aver-
age 61 cm VEV per ET is smaller in magnitude than at other high- Discussion
tonnage SAD ports (Mobile, Alabama and Tampa, Florida). This
suggests that the large total 61 cm VEVs at Savannah, Georgia, While VEV metrics describe the demands waterway users place on
are the result of many smaller encroachments rather than fewer the available channel depth, they do not fully describe waterway
large magnitude encroachments. This was also the case (albeit on management economics. To inform funding allocations, depth de-
a smaller scale) for Brunswick, Georgia. Brunswick had a relatively mand must be weighed against the effort required to supply the
large 61 cm VEV for its size and had relatively large numbers of channel dimensions. This level of effort can be expressed either
ETs—by the ratio of ET to total transits, Brunswick, Georgia, through the volume of sediment removed by maintenance dredging
ranks the highest of all ports in the study. From a usage perspective, of a port’s navigation channels or the associated costs in dollars. To
large numbers of vessels slightly encroaching on the bed are indic- account for this, the volume of sediment removed and associated
ative of the efficient use of available depth in the navigation chan- costs from 2015 to 2018 at the 13 studied ports were obtained
nel, as large numbers of users demand close to the full channel from the USACE Dredge Information System (DIS, USACE-DIS
depth, but the encroachment is minor on a per-vessel basis. In con- 2022). Table 3 displays the average 61 cm VEV, the average num-
trast, Mobile, Alabama has an average annual 61 cm VEV near that ber of 61 cm ETs, the average annual dredged volume, the average
of Savannah, Georgia (4.47 versus 5.50 Mm3 X T) despite far annual dredging expenditures, the ratio of dredge volume to
Fig. 5. (Color) Channel cross section displayed in an aerial view in Fig. 4. The shaded red region represents the 61 cm VEV, and the blue +
red-shaded region represents the 1.52 m VEV.
Table 3. Average 61 cm VEV (Mm3 X T), average 61 cm ET, average annual dredge volume (Mm3), average annual dredging expenditures (in millions),
ratio of average dredge volume to average 61 cm VEV, and ratio of average dredge cost per average 61 cm ET ($1,000/ET)
Avg. 61 cm VEV Avg. 61 cm Avg. dredge Avg. annual cost Avg. dredge vol./ Avg. cost/Avg. 61 cm
Port (Mm3 X T) ET (#) vol. (Mm3) ($ million) Avg. 61 cm VEV ET ($1,000/ET)
Mobile, Alabama 4.47 125 3.5 $7.80 0.78 $62.40
Savannah, Georgia and 6.15 600 5.7 $34.20 0.92 $57.00
Brunswick, Georgiaa
Tampa, Florida 1.09 59 0.4 $10.30 0.37 $174.58
Pascagoula, Mississippi 0.01 6 0.3 $1.40 35 $233.33
Port Everglades, Floridab 0 3 0 $0.00 N/A $0.00
Charleston, South Carolina 0.35 31 1.5 $9.40 4.35 $303.23
Jacksonville, Florida 0.06 14 0.5 $16.20 6.88 $1,157.14
San Juan, Puerto Rico 0.33 28 0.2 $1.70 0.35 $60.71
Miami, Floridab 0 3 0 $0.00 N/A $0.00
Wilmington, North Carolina 0.32 26 3.0 $21.60 9.23 $830.77
Canaveral, Florida 0.02 3 0.2 $3.30 12.5 $1,100.00
Palm Beach, Florida 0.01 2 0.1 $3.20 12.5 $1,600.00
a
Savannah, Georgia, and Brunswick, Georgia, are combined in DIS—the avg. 61 cm VEV and ET values are summed here for comparison.
b
No dredging is recorded in DIS for 2015–2018.
average 61 cm VEV, and the ratio of dredge expenditure to the av- volume to 61 cm VEV. Incorporating dredging costs captures the
erage number of ETs for each port. information necessary for decision makers to evaluate whether
USACE waterway management traditionally weighs the level of the cost per encroaching transit at Palm Beach, Florida ($1.6M/
effort required to supply the existing channel dimensions (either ET) is justified relative to the cost at Miami, Florida, or Port Ever-
through sediment volume removed or dredging expense) against glades, Florida (each $0), or with other high- (e.g., Mobile, Ala-
the throughput tonnage of the associated port. VEV-related metrics bama, $62.4 K/ET) to midtonnage ports (e.g., San Juan, Puerto
allow consideration of alternative perspectives incorporating vessel Rico, $60.71 K/ET).
demand for the full channel dimensions (i.e., available depth). One Regardless of the metric used to quantify the effort required to
alternative highlighted in Table 3 is to normalize the volume of supply the existing channel dimensions (dredge volume or ex-
sediment removed by the impact of accumulated sediment on tran- pense), the novel aspect of this approach is the ability to weigh
siting vessels as measured by the average 61 cm VEV. Smaller val- against the quantifiable user demand for the deepest parts of the
ues of this ratio indicate greater user demand for the deepest 61 cm channels. Examining the VEV-related metrics at CBDO = 61 cm
of the navigation channels relative to the volume of sediment re- identifies ports where vessels are more frequently observed to tran-
moved. Conversely, large values of this ratio indicate large vol- sit the channel with elevated risk underkeel clearance. Based on the
umes of sediment removed relative to the observed demand for observed demand, the USACE can determine which ports’ user
full channel depth. Wilmington, North Carolina and Charleston, base would benefit from additional dredging investment. However,
South Carolina are distinct among the large ratio ports, as they the resources available to maintain channels are finite. Therefore,
have roughly an order of magnitude larger average 61 cm VEVs the natural follow-up question is “Which ports’ users do not de-
than the other ports with ratios greater than unity. Therefore, mand the full extent of the channel depth and what would the im-
their larger ratios are due to relatively larger volumes of sediment pact of a reduction in dredging at those ports be?” In this respect,
removed rather than smaller average 61 cm VEVs. relying on 61 cm VEV metrics is insufficient, as they do not con-
A second alternative perspective is to consider the USACE as sider vessels utilizing the large majority of the available channel
spending dredging resources to obtain safe passage of deep drafting depth but maintaining lower risk clearance margins (i.e.,
vessels. For large commercial vessels, any encounter between the >61 cm). To remedy this, this study uses the VEV calculations
vessel’s keel and the channel bottom can result in adverse out- for CBDO = 1.52 m, specifically the difference between the 61 cm
comes. These range in severity and may include costs to vessel and 1.52 m VEVs (dVEV = 1.52 m VEV − 61 cm VEV). dVEV
owners, traffic delays, and major environmental catastrophes. represents the volume of the vessel hull that encroaches on the
Thus, the USACE could consider the ratio of the average dredging 1.52 m bed domain excluding the portion that encroaches on the
expenditure at a port relative to the number of average 61 cm ETs, 61 cm bed domain. Vessels encroaching on the 1.52 m bed domain
also displayed in Table 3. Smaller values of this ratio indicate that but not the 61 cm bed domain maintain low-risk clearance margins
more vessels are transiting within 61 cm of the bed (i.e., at higher at extant channel conditions but are sufficiently close to the bed that
risk) relative to the dollars spent maintaining the channel (note this they would have been adversely impacted by a reduction in dredg-
ratio interpretation does not apply to ports with zero expenditure, ing. dVEV allows the estimation of the extent of the adverse impact
such as Miami, Florida, and Port Everglades, Florida). Comparing for each port and the identification of those where impacts would be
dollars spent rather than volume dredged includes effects related to minimized.
the unit cost of dredging at each port (Frittelli 2019). For example, Table 4 displays the average dVEV, the average 61 cm VEV,
Tampa, Florida has a larger average unit cost than higher-tonnage and average annual dredging expenditures per port. Each port’s
ports in the SAD (e.g., Mobile, Alabama) and consequently has the rank among the studied ports by the average dVEV and average
fourth smallest ratio of average dredging expenditure per 61 cm ET 61 cm VEV appears in parentheses after the respective value. It
(174.58 $1,000/ET) despite the fact that it had the second smallest highlights ports where vessels were observed to transit with low
ratio of dredged volume to 61 cm VEV. Conversely, relatively risk (i.e., low ranking by average 61 cm VEV) and adverse impacts
cheap unit cost causes Pascagoula, Mississippi, to have the fifth due to potential reduction in dredging would have been minimized
smallest ratio of average dredging expenditure per 61 cm ET (i.e., low ranking by dVEV). Palm Beach, Florida (13th by dVEV,
(233.33 $1,000/ET) despite having the largest ratio of dredged 11th by 61 cm VEV); Miami, Florida (12th by dVEV and 61 cm
Table 4. Average dVEV (Mm3 X T), average 61 cm VEV (Mm3 X T), and average annual dredging expenditure ($ million)
Port Avg. dVEV Mm3 X T (rank) Avg. 61 cm VEV Mm3 X T (rank) Avg. annual cost ($ million)
Mobile, Alabama 44.50 (1) 4.47 (2) $7.80
Savannah, Georgia 29.18 (2) 5.50 (1) $34.20a
Tampa, Florida 7.78 (3) 1.09 (3) $10.30
Pascagoula, Mississippi 0.18 (10) 0.01 (10) $1.40
Port Everglades, Floridab 0.11 (11) 0 (13) $0.00
Charleston, South Carolina 1.82 (7) 0.35 (5) $9.40
Jacksonville, Florida 1.66 (8) 0.06 (8) $16.20
San Juan, Puerto Rico 2.80 (4) 0.33 (6) $1.70
Miami, Floridab 0.07 (12) 0 (12) $0.00
Wilmington, North Carolina 2.48 (5) 0.32 (7) $21.60
Canaveral, Florida 0.27 (9) 0.02 (9) $3.30
Brunswick, Georgia 2.29 (6) 0.66 (4) $34.20a
Palm Beach, Florida 0.02 (13) 0.01 (11) $3.20
Note: Each port ranked by dVEV and 61 cm VEV in parentheses.
a
Savannah, Georgia, and Brunswick, Georgia, are combined in DIS–this average annual cost represents the sum for both.
b
No dredging expenditure is recorded in DIS for 2015–2018.
Table 5. Average 61 cm VEV (Mm3 X T), average 61 cm ET, average annual dredge volume (Mm3), average annual dredging expenditures ($ million), ratio
of average dredge volume to average 61 cm VEV, and ratio of average dredge cost per average 61 cm ET ($1,000/ET) for Southwest Pass, Louisiana
Avg. 61 cm VEV Avg. 61 cm Avg. dredge vol. Avg. annual cost Avg. dredge vol./Avg. Avg. cost/Avg. 61 cm ET
Port (Mm3 X T) ET (#) (Mm3) ($ million) 61 cm VEV ($1,000/ET)
Southwest Pass, 0.66 113 42.82b $131.00b 65.12 $1,159.29
Louisianaa
a
Does not include 2018.
b
Actual/projected dredge volumes and actual/projected costs found in USACE (2017) were used.
VEV); Port Everglades, Florida (11th by dVEV, 13th by 61 cm Reanalysis of Southwest Pass, Louisiana
VEV); Pascagoula, Mississippi (10th by dVEV and 61 cm VEV);
The Southwest Pass, Louisiana, data presented in Scully and
and Canaveral, Florida (9th by dVEV and 61 cm VEV) meet
Young (2021) was reanalyzed with the methodology outlined
these criteria. However, the observed dredging expenditures at
here to compare the studied SAD ports with a navigation channel
Port Everglades, Florida, and Miami, Florida ($0), indicate that
supporting extremely high cargo tonnage. Southwest Pass,
these two ports require little maintenance. Consequently, while
Louisiana averaged 265 million tons over the 2015–2017 time pe-
these low values of the 61 cm VEV and dVEV translate to ineffi-
riod covered in the Scully and Young (2021) data set, exceeding the
cient use of available depth on the part of the users of Port
combined average annual tonnage of all 13 studied SAD ports by
Everglades, Florida, and Miami, Florida, navigation channels,
6 million tons. Tables 5 and 6 display the results for Southwest
from the perspective of waterway management, they are highly ef-
Pass, Louisiana, specifically the same metrics found in Tables 3
ficient ports regardless. In contrast, Palm Beach, Florida;
and 4.
Pascagoula, Mississippi; and Canaveral, Florida collectively cost
Despite its high tonnage, Southwest Pass, Louisiana, has a com-
$7.9 million to dredge on average; however, their user base on av-
paratively low average 61 cm VEV (0.66 Mm3 X T), less than that
erage encroaches on the 1.52 m bed domain nearly 10 times less
of Jacksonville, Florida (which ranked eighth by average 61 cm
than the Mobile, Alabama, user base encroaches at 61 cm
VEV within the studied SAD ports). The average annual dredge
(1.52 m VEV of 0.49 Mm3 X T versus 61 cm VEV of 4.47). Effec-
volume was also nearly an order of magnitude larger than the clos-
tively, the USACE expends similar levels of resources ($7.9 mil-
est SAD port (42.82 Mm3 versus 6.15 for Savannah, Georgia/
lion for Palm Beach/Pascagoula/Canaveral versus $7.8 million
Brunswick, Georgia). Consequently, its ratio of average dredge
for Mobile) to meet one-tenth the demand for the bottom 1.52 m
volume to average 61 cm VEV is nearly double the closest SAD
of the available channel depth at these three ports as it does to
port (65.12 versus 35.00 for Pascagoula, Mississippi). Southwest
meet the higher risk demand of vessels for the bottom 61 cm of
Pass, Louisiana, costs $131.00 million per year to dredge on aver-
the available channel depth at Mobile, Alabama.
age, exceeding the combined average annual spending for all 13
studied SAD ports by $21.90 million. As a result, Southwest
Table 6. Average dVEV (Mm3 X T), average 61 cm VEV (Mm3 X T), and
Pass, Louisiana, has an average cost per 61 cm ET higher than
average annual dredging expenditures ($ million) for Southwest Pass,
Louisiana
any port in SAD apart from Palm Beach, Florida ($1.2 million/
ET versus $1.6 million/ET at Palm Beach, Florida). The average
Avg. annual annual dVEV for Southwest Pass, Louisiana, is comparatively
Avg. dVEV Avg. 61 cm VEV cost larger relative to the studied SAD ports, 18.11 Mm3 X T, but
Port (Mm3 X T) (Mm3 X T) ($ million)
would still rank behind Mobile, Alabama, and Savannah, Georgia.
Southwest Pass, 18.11 0.66 $131.00b Thus, the adverse impacts on observed vessel transits due to the po-
Louisianaa tential reduction in dredging at Southwest Pass, Louisiana, are non-
a
Does not include 2018. trivial, though not larger than Mobile, Alabama or Savannah,
b
Actual/projected dredge volumes and actual/projected costs found in Georgia, despite Southwest Pass’s, Louisiana, higher average ton-
USACE (2017) were used. nage. This analysis is meaningful because dredges have been
rerouted from active projects in other USACE districts to relieve user demand for channel depth in the United States, as large vol-
Southwest Pass, Louisiana shoaling recently (Hartman et al. umes of cargo are transported on shallow-drafting vessels at many
2022) at great expense to the USACE and the original project ports (USACE-IWR 2012).
schedules. Using VEV-related metrics, the USACE can weigh Of the studied ports, the top three by tonnage (Mobile, Alabama;
the implications for trafficking vessels when making active dredge Savannah, Georgia; and Tampa, Florida, respectively) consistently
redeployment decisions in the future. rank highly by the 61 cm VEV metrics described here: average an-
nual 61 cm VEV, average number of 61 cm encroaching transits,
and average encroaching-transit-normalized 61 cm VEV. The VEV-
Policy Implications
related metrics at a bed domain offset of 61 cm identify ports where
Underkeel clearance-related metrics for vessel transits, such as vessels are more frequently observed to transit the channel with
VEV, quantify the extent to which each vessel makes use of higher-risk underkeel clearance and may benefit from additional
the USACE depth-provision service. The applicability of VEV dredging investment. That such large ports may benefit is not surpris-
to waterway maintenance prioritization is demonstrated previ- ing as they are likely to have the most vessel transits, which provides
ously, but this capability could have broader ramifications for more opportunities for vessel encroachment. To determine in which
addressing the constitutional questions and international percep- ports users do not require the full extent of the extant channel depth,
tion of the present structure of the Harbor Maintenance Trust this study proposes the difference between the 1.52 m and 61 cm
Fund (HMTF) and associated fee assessment. VEV directly VEVs (dVEV). The dVEV metric accounts for ports where vessels
links individual waterway users and the environmental condi- transit with low-risk clearance margins at the extant level of dredging
tions (including USACE-maintained depth) at the time of transit. but are sufficiently close to the bed where they would be adversely
An HM fee based on vessel clearance addresses the key defi- impacted by a reduction in dredging. Based on the low 61 cm
ciency identified by the US Supreme Court in United States v. VEV and dVEV metrics, as well as the dredge funding allocations
United States Shoe Corp. 532 U.S. 360 (1998), which held over the study period, Palm Beach, Florida; Pascagoula, Mississippi;
that the HMTF fee was a de facto tax violating the Export Clause and Canaveral, Florida would be least impacted by a reduction in
of the Constitution and did not apply to exporters. The US Su- maintenance dredging.
preme Court noted in the same decision that the Export Clause VEV represents a new data-driven methodology to quantify
does not “rule out a ‘user-fee’”; that is “a charge designed as channel maintenance requirements based on measured usage by
compensation for government-supplied services, facilities, or transiting vessels, allowing waterway managers to consider
benefits” [Pace v. Burgess (1875)]. However, the “connection
dredge funding and active dredge redeployment decisions
between a service the government renders and the compensation
through the lens of supply (of available depth) and demand (for
it receives for that service” was not sufficiently close for a user
that depth by users). At the local port level, VEV and related
fee as the HM fee is currently assessed (based on cargo value).
underkeel clearance metrics provide USACE District personnel
This also addresses concerns by US international trade partners
with independent data-driven information on vessel usage of
that the HM fee represents a tariff on imported products (rather
available depth in their navigation channels. As the methodology
than a usage fee) and violates US obligations under the World
Trade Organization (Fritelli 2011). Similarly, the aggregate an- is scaled up across the USACE portfolio, additional features may
nual VEV values for each port could inform HMTF disburse- be desirable to include. Although the selection of 61 cm of bed
ments, alleviating other concerns raised in the study by Fritelli domain offset to demarcate high-risk transits follows general
(2011). Specifically, (1) certain channels that receive little-to-no PIANC guidance for vessels drafting ≤12.19 m, vessels carrying
deep draft vessel traffic (and pay little-to-nothing into the specific commodity types [e.g., liquefied natural gas (LNG) or pe-
HMTF) receive an outsized percentage of disbursements; and troleum tankers] may require different levels of clearance to
(2) vessels with high cargo value but lower draft (frequently con- maintain appropriate safety margins. Similarly, the selection of
tainer vessels) pay a disproportionate amount into the HMTF rel- 1.52 m of bed domain offset to identify vessels at risk of disrup-
ative to their channel usage. tion due to a reduction in dredge maintenance could be modified
to incorporate the USACE Corps Shoaling Analysis Tool’s
(CSAT—Dunkin et al. 2018) shoaling rates for each respective
Conclusions and Future Work channel to better approximate the expected change in the bed
over a specified temporal duration in the absence of dredging.
This study presents metrics related to the clearance between the These effects could be incorporated into the processing steps by
keels of transiting vessels and the channel bed, specifically vessel varying the higher-risk bed domain offset (61 cm) based on the
encroachment volume [VEV, Eqs. (1) and (2)], to quantify the AIS-reported vessel type and the lower-risk bed domain offset
user demand for the deepest portions of the navigation channels (1.52 m) based on channel-specific shoaling rates without altering
for 13 SAD ports managed by the USACE. The specific bed do- the proposed approach. Additionally, crowdsourced bathymetry
main offsets used to compute the VEV (61 cm versus 1.52 m) measurements could be incorporated into the methodology to pro-
allow this demand to be segregated by the risk of disrupted tran- vide more temporal resolution to the available channel depth mea-
sits. The metrics are explored to demonstrate the feasibility of up- surements. The CSAT shoaling rates could also be used to
dates to the USACE dredge maintenance funding rubric. The interpolate between USACE channel surveys, accomplishing
USACE currently prioritizes maintenance funding allocations the same goal of improving the temporal resolution of the avail-
based on the criticality of a port to nationwide commerce (as es- able depth. This improvement would be particularly impactful
timated by tonnage) and by channel condition (as estimated by in areas where channel surveys are conducted less frequently. Fi-
channel depth relative to authorized depth). This approach does nally, in certain channels, the horizontal clearance experienced by
not link the supply of the service provided by the USACE (i.e., vessels (the distance between the vessel and the channel edge) is
channel depth and width) to the users’ demand for that service of greater concern than the underkeel clearance. The horizontal
(i.e., how much of the available depth navigating vessels use). clearance calculation could be integrated into the workflow de-
Tonnage has been recognized as an imperfect approximation of scribed here and reported as an additional metric.
Data Availability Statement PIANC (World Association for Waterborne Transport Infrastructure).
2014. Harbour approach channels—Design guidelines. PIANC Rep.
Some or all data, models, or codes generated or used during the No. 21. Brussels, Belgium: PIANC.
study are available in a repository or online in accordance with PIANC (World Association for Waterborne Transport Infrastructure).
2020. Resilience of the maritime and inland waterborne transport sys-
funder data retention policies: Marine Cadastre AIS data (Bureau
tem. PIANC Environmental Committee Task Group 193. Brussels,
of Ocean Energy Management and National Oceanic and Atmo- Belgium: PIANC.
spheric Administration (BOEM and NOAA 2019)), eHydro sur- Rawson, K. J., and E. C. Tupper. 2001. Basic ship theory. 5th ed. Oxford:
veys (USACE 2020b), tidal elevation (NOAA 2020), vessel draft Butterworth-Heinemann.
(USACE-IWR 2018), and dredging records (USACE-DIS 2022). Scully, B. M., and K. N. Mitchell. 2017. “Underkeel clearance reliability
model for dredged navigation channels.” Transp. Res. Rec. 2611: 41–
49. https://doi.org/10.3141/2611-05.
Scully, B. M., and D. L. Young. 2021. “Evaluating the underkeel clearance
of historic vessel transits in the Southwest Pass of the Mississippi river.”
J. Waterw. Port Coastal Ocean Eng. 147 (5): 05021008. https://doi.org
/10.1061/(ASCE)WW.1943-5460.0000655.
Scully, B. M., D. L. Young, and J. E. Ross. 2020. “Mining marine vessel
AIS data to inform coastal structure management.” J. Waterw. Port
Coastal Ocean Eng. 146 (2): 04019042. https://doi.org/10.1061
/(ASCE)WW.1943-5460.0000550.
Szlapczynski, R., and J. Szlapczynska. 2016. “An analysis of domain-based
ship collision risk parameters.” Ocean Eng. 126: 47–56. https://doi.org
References /10.1016/j.oceaneng.2016.08.030.
Szlapczynski, R., and J. Szlapczynska. 2021. “A ship domain-based model
BOEM and NOAA (Bureau of Ocean Energy Management and National of collision risk for near-miss detection and collision alert systems.”
Reliab. Eng. Syst. Saf. 214: 107766. https://doi.org/10.1016/j.ress
Oceanic and Atmospheric Administration). 2019. “Automatic identifi- .2021.107766.
cation system data.” Accessed January 23, 2019. https:// USACE. 2006. Hydraulic design of deep-draft navigation projects. EM
marinecadastre.gov/ais. 1110-2-1613. Washington, DC: USACE.
Byrnes, M. R., J. L. Baker, and F. Li. 2002. Quantifying potential measure- USACE. 2017. “Mississippi river ship channel Gulf to Baton, LA final in-
ment errors and uncertainties associated with bathymetric change tegrated general reevaluation report and supplemental environmental
analysis. ERDC/CHL CHETN-IV-50. Vicksburg, MS: Coastal and impact statement Appendix K – preliminary assessment of the dredge
Hydraulics Laboratory, USACE. material management plan.” Accessed December 22, 2022. https://www
Curtis, B. 2018. “An integrated approach to port planning, operations, and .mvn.usace.army.mil/Portals/56/docs/Projects/Miss%20Deep/13_Append
risk management through technology.” In Proc., 34th PIANC World ix%20K_Dredge%20Material%20Management%20Plan.pdf.
Congress, 1–17. Brussels, Belgium: World Association of USACE. 2020a. Civil works direct program development policy guidance.
Waterborne Transport Infrastructure (PIANC). Washington, DC: USACE.
Dunkin, L. M., L. A. Coe, and J. J. Ratcliff. 2018. Corps shoaling analysis USACE. 2020b. USACE hydrographic surveys. Washington, DC: USACE.
tool: Predicting channel shoaling. ERDC/CHL TR-18-16. Vicksburg, USACE. 2022. Charleston entrance channel. CCR Rep. No. 4020
MS: Coastal and Hydraulics Laboratory, USACE. 20220328. Washington, DC: USACE.
Fritelli, J. 2011. “Harbor maintenance trust fund expenditures. USACE-CPT (Channel Portfolio Tool). 2022. Channel portfolio tool.
Congressional research service.” Library of Congress. Washington, Washington, DC: USACE.
DC: Congressional Research Service. USACE-DIS (Dredge Information System). 2022. Dredge information sys-
Frittelli, J. 2019. Harbor dredging: Issues and historical funding. tem webtool. Washington, DC: USACE.
Washington, DC: Congressional Research Service. USACE-IWR (Institute for Water Resources). 2012. U.S. port and inland
Garrido, J., S. Saurí, Á. Marrero, Ü. Gül, and C. Rúa. 2020. “Predicting the waterways modernization: Preparing for post-panamax vessels.
future capacity and dimensions of container ships.” Transp. Res. Rec. Washington, DC: USACE.
2674 (9): 177–190. https://doi.org/10.1177/0361198120927395. USACE-IWR (Institute for Water Resources). 2018. Waterborne foreign
Hartman, M. A., K. N. Mitchell, L. M. Dunkin, J. Lewis, B. Emery, N. F. cargo. Washington, DC: USACE.
Lenssen, and R. Copeland. 2022. “Southwest pass sedimentation and USACE-WCSC (Waterborne Commerce Statistics Center). 2021. Ports
dredging data analysis.” J. Waterw. Port Coastal Ocean Eng. 148 (2): and waterways webtool. Washington, DC: USACE.
05021017. https://doi.org/10.1061/(ASCE)WW.1943-5460.0000684. USCG (US Coast Guard). 2018. “Maritime commerce strategic outlook.”
Mahar, I. D. 2018. “Dynamic under keel clearance project.” Coast Guard Accessed July 21, 2021. https://www.dco.uscg.mil/Our-Organization/
J. Saf. Secur. Sea 75 (2): 30–32. Assistant-Commandant-for-Prevention-Policy-CG-5P/Marine-Transpo
Mitchell, K. W. 2009. “Depth-utilization analysis for estimating economic rtation-Systems-CG-5PW/Maritime-Commerce/.
activity supported by dredging.” Terra Aqua 116: 22–30. Waters, J. K., R. H. Mayer, Jr., and D. L. Kriebel. 2003. “Dredging-related
NOAA (National Oceanic and Atmospheric Administration). 2020. Tides implications of current and future trends in the international shipping
and great lakes water levels—NOAA tides and currents. Washington, industry.” In Dredging’02: Key Technologies for Global Prosperity,
DC: NOAA. edited by S. Garbaciak, 1–14. Reston, VA: ASCE.
OMC International. 2022. “Dynamic under keel clearance.” Accessed Young, D. L., and B. M. Scully. 2018. “Assessing structure sheltering via
October 31, 2022. https://omcinternational.com/products/dukc/. statistical analysis of AIS data.” J. Waterw. Port Coastal Ocean Eng.
Parker, B. B., and L. C. Huff. 1998. “Modern under-keel clearance manage- 144 (3): 04018002. https://doi.org/10.1061/(ASCE)WW.1943-5460
ment.” Int. Hydrogr. Rev. 75 (2): 143–165. .0000445.
REPORT DOCUMENTATION PAGE
1. REPORT DATE 2. REPORT TYPE 3. DATES COVERED
December 2023 Final Report START DATE END DATE
5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER
6. AUTHOR(S)
David L. Young, Brandan M. Scully, Sean P. McGill, Ashley J. Elkins, and Marin M. Kress
14. ABSTRACT
The US Army Corps of Engineers (USACE) traditionally uses two metrics to evaluate the maintenance of coastal navigation projects:
tonnage at the associated port (representing relative importance) and the controlling depth in the channel (representing operating
condition). These are incorporated into a risk-based decision framework directing funds where channel conditions have deteriorated and
the disrupted tonnage potential is the highest. However, these metrics fail to capture shipper demand for the maintained depth service
provided by the USACE through dredging. Using automatic identification system (AIS) data, the USACE is pioneering new metrics
describing vessel demand for the channel depth, represented by vessel encroachment volume (VEV). VEV describes the volume of the
hull intruding into a specified clearance margin above the bed and captures how much vessels use the deepest portions of USACE-dredged
channels. This study compares the VEV among 13 ports over 4 years by combining AIS, tidal elevations, channel surveys, and sailing
draft. The ports are ranked based on the services demanded by their user base to inform the decision framework driving dredge funding
allocations. Integrating demand for-depth metrics into the Harbor Maintenance Fee assessment and/or Trust Fund disbursements could
alleviate the constitutionality concerns and several criticisms levied against Harbor Maintenance funding.
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT 18. NUMBER OF PAGES
a. REPORT b. ABSTRACT C. THIS PAGE SAR 17
Unclassified Unclassified Unclassified
19a. NAME OF RESPONSIBLE PERSON 19b. TELEPHONE NUMBER (include area code)