Technical ReportPDF Available

AdjustaDepth Phase 1 Final Report DE-AR0000916 Feed the world on the way to abundant biofuels

  • OceanForesters


AdjustaDepth Project was funded by the U.S. Department of Energy Advanced Research Projects Agency - Energy to design an artificial reef system for growing and harvesting seaweed for advanced sustainable biofuels. The team discovered that the system grows more seaweed per hectare when it is part of a complete ecosystem with shellfish, finfish, and other animals, hence it can "feed the world on the way to abundant biofuels."
Team Members: 1
Primary Technical Targets: 2
Design descriptions, Structure: 3 – 7
Design descriptions, Biology 8 – 12
Capital: 13
Operations: 14 21
Mechanical resiliency confirmation: 22 29
Equipment use efficiency: 30
Deployment area rationale: 31
Scalability: 32 33
Nutrient requirements and provision: 34 -37
Biomass production cost: 38 – 40
Energy return: 41
Critical assumptions: 42 44
Sensitivity analysis: 45 46
Path-to-Market: 47 – 50
AdjustaDepth Phase 1 Final Report DE-AR0000916
Table of Contents
Impact: Replace all fossil fuels
Feed the world on the way to abundant biofuels
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Team Members
Dr. LaPointe 1990
Coordination, Permitting, Techno-Economic
ØKelly Lucas, University of Southern Mississippi
ØReginald Blaylock, University of Southern Mississippi
ØStephan Howden, University of Southern Mississippi
ØMark Capron, OceanForesters,
ØJim Stewart, OceanForesters
Biology and Nutrients
ØMichael Chambers, University of New Hampshire
ØScott James, Baylor University
ØMaureen Brooks, University of Maryland
ØStacy Krueger-Hadfield, University of Alabama
ØSuzanne Fredericq, University of Louisiana at
ØBrian Lapointe, Florida Atlantic University
ØAntoine N'Yeurt, University of the South Pacific
ØRicardo Radulovich, University of Costa Rica
ØAlejandro Buschmann-Rubio, University of Lagos,
Oceanography and Technology
ØSteven DiMarco, Texas A&M University,
ØKerri Whilden, Texas A&M University,
ØTo ny Kna p, Te xa s A &M Un iv ers ity,
ØMH Kim, Texas A&M University
ØChris Webb, AI Control Technologies Inc,
ØAlberto Mestas-Nunez, University of Texas,
San Antonio
Structural Modeling
ØRob Swift, University of New Hampshire,
ØZach Moscicki, University of New Hampshire
ØIgor Tsukrov, University of New Hampshire,
ØDavid Fredriksson, U.S. Naval Academy,
ØTo by Dew hur st , M ain e M ari ne Co mpo si tes ,
ØAndrew Drach, Callentis Consulting Group
Technology Advisors
ØSamson Rope
ØApplied Fiber
ØAquuai Robofish
ØSynthetik-Te c hn ol o gi es
AdjustaDepth Phase 1 Final Report DE-AR0000916
Primary Technical Targets
TEA Technical Target Results
Full System
TEA numbers are based on 1,000
-ha planted with optimum density.
Seafloor footprint would be 6,000
Range of
TEA range is
1 million-planted-ha, 6 million-ha footprint between 50 to
-m seafloor depth in the Gulf of Mexico. TEA structure cost based
on the average (75
-m) seafloor depth. Robust structure applicable to
300 million ha of U.S. continental shelf.
As low as
$50/DMT when other reef products pay 75% of structure and
monitoring costs and production is the same every day, year
-round. As
low as
$120/DMT without other reef products, but $290/DMT with ‘all
middle’ assumptions.
Net Energy
40:1 with room for increase. Nearly all the energy use is in mow-
transport to shore. Drops to near 20:1 when considering energy
embedded in structure.
Could jump to 100:1.
Infinitely scalable
. The first few farms would start with excess
nutrients from the Mississippi River. When the CleanCarbon
process comes on
-line (with biosolids and other biomass ‘waste’), the
-harvest-transport system design is ideal for back-haul and
distribution of nutrients while mowing
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 1 -Spiderweb structure1 Isometric
The AdjustaDepth structure shown here is designed to withstand extreme waves and currents characteristic of the
Gulf of Mexico. The spoked circular pattern can accommodate changing wind, wave and current directions present
during hurricanes. Netting substrate for the cultivation of Gracilaria tikvahiae spans the spokelines. Structural
ropes delineate netting panels and provide strength and tension across the netting membrane. Oyster cultivation
systems are hung from the perimeter where the oysters will filter incoming microalgae-laden water, and produce
nutrient rich waste for the Gracilaria to consume. An innovative anchor line design will maintain tension on the
system regardless of whether the structure is submerged or surfaced.
500 meters
20-ha, 50 acres
Table of ContentsBack to
The AdjustaDepth project began with basic
evaluation and weighing of different design
concepts. Pre-kickoff concepts featured a rigid
circumferential ring with adjustable buoyancy
(below center). Internal parallel tension members
supported algae growth substrate (bottom left).
During the MARINER kick-off team design
evaluations revealed that long rigid structural
components would not suffice at the desired
scales. More flexible, rope-based structures were
explored. Several adjustable depth concepts were
considered (bottom right). We settled on a circular
radial structure with radial spokes (top right).
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 2 Preliminary Design Concept
Table of ContentsBack to
Preliminary Quasi-Static load analyses yielded estimates for required
structural strengths, weights and buoyancies. As biological, operational,
construction constraints were investigated and established, the geometry of
the design details evolved. Different anchor line specifications were
investigated for their effect on system performance: rope length and elasticity,
chain length, submerged float size and position and ballast position.
Ultimately, structural performance observed in the dynamic modelling results
informed the present multi-stage anchor line design (third down from top
right) which should result in superior tension maintenance and depth control
on all anchor lines in all operating conditions. In phase 2 we intend to further
iterate the anchor line design for optimal tension management, targeting a
minimum anchor line tension of 10 kN in all but the most extreme conditions
Parallel design efforts focused on algae substrate as we considered methods
to reduce risk of marine mammal entanglement, cable termination options,
and tension management. Throughout Phase 1 the substrate structure
evolved into a design that is more robust and easily deployed. (Lower
figures, left to right, regulator-required 10x10-cm square grid not shown.)
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 3 Structural Design Progression
Table of ContentsBack to
The Phase 2 system improves tension maintenance by placing
the adjustable ballast on the anchor line.
All the Phase 1 dynamic analysis runs
assumed this anchor line geometry
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 4 -Spiderweb structure2 Plan View
Table of ContentsBack to
Screw/helix anchors are not well-
suited for the expected horizontal
loading of the anchors in the
AdjustaDepth design and are
expensive. Drag anchors are more
suited to the horizontally dominated
extreme loads we expect and more
cost effective.
Operation and maintenance vessels will be able
to moor to any float. Even when submerged, the
loads from a Cat 5 hurricane greatly exceed
those imposed by mooring a maintenance vessel
during surfaced weather conditions.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 5 -Spiderweb structure3 Profile, both positions
This figure shows the improved tension management system. The structure can be remotely submerged to
where wave forces are reduced, allowing for a less costly mooring system and little, if any, crop damage. Early
hydrodynamic forcing estimates indicated that submerging the system 25 meters below the ocean surface can
cut structural loads on the system by more than 50%. The structure and submerging system have evolved
such that we can select the submergence depth. All components can be kept 30 to 40-m below the ocean
surface in 50-m water depth, deeper when the seafloor is deeper. (Possible exception for the downstream
submersible floats, which might rise within about 20-m of the ocean surface during the 50-yr storm event.)
Table of Contents Back to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 6 Gracilaria tikvahiae design1 Biology-Structure compatibility
Gracilaria tikvahiae (Gtik) grows to the shape, size, and appearance
of cheerleader pom-poms. See picture at left. The long-life structure
design complements how Gtik grows:
§Individual Gtik plants live up to 40 years (Engel 2001).
§The structure and square net (10x10-cm) substrate forms a
relatively smooth plane, appropriate for tight (couple centimeters)
tolerances while floating/rolling an automated mowing device
over the plane.
§Gtik does not have air bladders (not expecting damage when the
structure submerges)
§Plant spacing and substrate on the structure is similar to how
Gtik is grown on rafts.
Dr. Lapointe 1990
0.0 2.0 4.0 6.0 8.0
Yield RATE -
Density -DMT/ha
Mow operating area
The structure is designed to support and harvest Gtik
maintained at the optimum density for yield. The black line in
the graph at right indicates the actual yield rates that produced
the total of 124 DMT/ha/yr Brian Lapointe measured with
frequent harvest of Gracilaria tikvahiae in 1978 and in 1999
(Capo, Lapointe et al. (1999) sustained a mean of 145
DMT/ha/yr over four years). Refer to the TEA ‘Yield-
The black line is an average of winter and summer yield rates.
It shows that maintaining relatively low density is one of the
keys to high yield. We will mow once a month in winter and
every two weeks in summer. Every time we mow, we harvest 5
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 7 Gracilaria tikvahiae design2 Yield-Structure compatibility
The graph on the previous slide indicates how our mowing
operation could achieve >100 DMT/ha/yr by maintaining
Gtik density between 3 to 8 DMT/ha (2-6 wet kg/m2).
The table above shows the annual yield that results from
either situation:
1) Supply just enough nutrients for 3.8%/day growth rate
and harvest the same amount (180 kg/ha/day) 365
days per year.
2) Semi-levelized nutrient supply for the 1978 Lapointe
growth rates and vary the harvest rate from 180
kg/ha/day in winter to 450 kg/ha/day in summer.
Table is ex cerpt from TEA ‘Yiel d -growthParameters’ Rows
One way annual yield is calculated in TEA
Yield from perennial system with frequent mowing and levelized growth rate
Lapointe 1978, yield with frequent mowing and semi-levelized growth rate
Levelized and semi-levelized %growth/day
when recycling nutrients
Based on La Pointe 1978 Level
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 8 Gracilaria tikvahiae design3 Planting-Structure compatibility
Self-seeding may be impractical or enhanced by green sea-turtle herbivory. The enhancement would be if sea turtles clean surfaces
such that more productive biomass grows. Sea turtles appear to detect a few m2of Gracilaria from tens of kilometers. Green sea
turtles are rare in the northern Gulf of Mexico, so not expected during Phase 2. However, we expect they will find full-scale
operations. At full-scale, turtles are not expected to eat more than 10% of the crop, which may not matter, if operating for levelized
The TEA is based on planting cuttings inside shields
initially and at as-needed intervals after that. Because
the structure is designed for long life, like a natural reef,
we can expect self-seeding, if we are careful to plant
male and female cuttings.
Fig. 1 (A) Diplontic life cycle.Fertilization (F)
rapidly follows meiosis (M), such as exhibited by
animals.Gametes are produced by meiosis.(B)
Haplodiplontic life cycle where meiosis and
fertilization are spatiotemporally separated.The
mature diploid produces haploid spores via
meiosis.These haploid spores settle and become
free-living, macroscopic, haploid individuals.
Haploids produce gametes, and after fertilization,
the diploid spores settle and become the next
generation of diploids.(C) Uncoupled
haplodiplontic life cycle when asexual
reproduction occurs.(adapted from Krueger-
Hadfield et al.2016)
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 9 Gracilaria tikvahiae design4 On-structure profiles
The illustrations at left are suggestive of the
profile (section-elevation, sideview) shapes
that frequently mowed Gtik may assume.
Male shapes may be different from female
shapes. Haploid individual shapes may be
different from diploid shapes.
The illustrations are roughly to scale for
mowing when plant density is near 8 wet
kg/m2(8 DMT/ha). Recently mowed
density would be near 3 wet kg/m2(3
When harvesting the same amount every
day of the year (levelized harvest), the time
between mowing would be steady year-
round, perhaps 30 days. The height of the
mow would be adjusted to collect the
desired biomass.
Note that nutrient distribution from the
mower (or along the conveyor) falls on at
least a third of the biomass those nutrients
will grow.
20-cm between herbivory shields
Full-grown Gtik plants
Ready to mow Gtik, 60 to 80% of full growth
Recently mowed Gtik
Netting substrate
Netting substrate
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Design descriptions 10 Gracilaria tikvahiae design5 Turtle-Gtik-Structure compatibility
Because both structure and Gtik are
long-lived, the plants might survive
herbivory if the cuttings or seeded
twine is placed in a shield, shown on
the Planting slide.
Dr. Radulovich’s experience suggests
the turtles will not eat the shield and
the protected portion of the plant will
grow back through the openings in the
tube. Therefore, we would not need
to replant areas eaten by turtles.
Dr. Suzanne L Fredericq’s research
suggests the Green sea turtle
(Chelonia midas) is not expected in
the Central Gulf during Phase2.
Dr. Ricardo Radulovich’s experience:
Sea turtles act as if they can detect a dense 20-m2patch of Gracilaria from many kilometers.
The turtles ate Gracilaria from his rafts from below .
Turtles begin eating up the softer, newer growth yet move into the thicker material, cleaning all traces of seaweed off
the rope fibers. They never damaged a rope nor did we see that a rope had been nibbled.
The turtle experience was in the Pacific (inside the Gulf of Nicoya in Costa Rica), not the Caribbean. Not sure what ate
the seaweed in the Caribbean.
One problem assessing the extent of herbivory is scale. If the trial is small (even a few thousand square meters) effects
will be magnified because a few herbivores can inflict lots of damage.
Turtles are capable of cutting through the low-density polyethylene sleeve mesh. Question is: will they do it?
Table of ContentsBack to
Nth-scale range of Green Sea Turtle
AdjustaDepth Phase 1 Final Report DE-AR0000916
Installed capital costs
All low All middle All high
Structure: anchors, ropes, floats, ballasts,
connections, netting, and installation
Mechanical (depth adjust), electrical (solar PV),
monitoring, and control on the structures
-harvest-transport units , rounded (See 'Mow-
Harvest$' Row 46 for variations)
Total initial capital, rounded
Capital costs for one 1,000-ha aggregate perennial, 15+year life, combination seaweed farm and flexible
floating reef 50 reefs, 20-ha each, about 6,000-ha seafloor footprint, built with the strength to support
filter feeders for nutrient conversions and materials appropriate for long-life.
All low The most possible but optimistic assumptions in every case for the lowest cost and most energy.
All middle A realistic assumption for every case.
All high The most pessimistic ‘surprise-it-that-bad’ assumption for every case.
Uncertain regulatory requirements on the substrate opening size drive the variation in structure cost.
Mechanical system cost varies primarily with the number of days for recharging the compressed air.
Mow-harvest-transport capital is proportional to yield, degree of yield/harvest leveling, nutrient
variations, and barge size. In the table above, several parameters cause the erratic change in the
capital cost of mow-harvest-transport units.
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations1 Overview
Shore to 1st
structure Mow-harvest 1st Structure to
2nd, 3rd, 4th
Structure to
Release air
Sink to near
Wait out
Add compressed
air to adjustable
Deploy structure
Plant Gtik, component
replacements: 1st, every 5 to 20
Replace structure,
15 to 20 years
Mow-harvest-transport Once a month per structure with year-round levelized production of 30 DMT/ha/yr.
Varies from eve ry 10 d ays to ever y 30 days for maximum annual yield. Recycled nutrients can ret urn on the backhaul.
Submerge to avoid storms 2 to 5 times per year. Key metric: Time between storms determines size
and capital cost for solar panels and air compressors.
Structure and planting Long-life Gtik plants on long-life structure, with component replacements, becomes
a permanent floating flexible reef.
Gang mower
Stationary barge
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations2 Pre-fabrication and deploying at Nth scale
Table of ContentsBack to
outer circumferential
rope, pre-cut on rolls
spokes, pre-
cut on rolls
inner circumferential rope
and/or bottom netting, pre-
cut on rolls
working face
The figure at left shows pre-fabrication of
the growing surface with the structural
spokes and outer circumferential ropes in
a rectangle that can be as small as 50-m
by 550-m.
Construction (making connections)
proceeds from north to south. Rolling up
the connected portion also proceeds from
north to south, lagging 10 to 20 m behind
the working face. If the structure is rolled
around a buoyant core, the core could
serve as floatation when towing.
Ideally the work area is such that the
rolled-up portion of structure is rolled
directly into the water, after first wrapping
in protective material.
The rolled material is estimated to weigh
30 MT, more with the core. See TEA
‘PreFab&Instl’ tab.
Plan view prior to rollup.
The anchors and chains would be set months in advance so that the ropes
arrive at the pre-fab site pre-cut and with Applied Fiber terminations. After
towing to site, the structure is unrolled. Spoke-floats and pre-cut anchor lines
with tension management are connected with specialized equipment.
Self-propelled gang
mower and nutrient
distributor spirals
around the structure
Temporarily moored bar ge
Seaweed and
nutrient conveyor
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations3 Mow-harvest-transport Large Barge (5,000 ton displacement)
Most of the energy use is in the
travel from shore to structures
(loaded with nutrients) and return
from structures to shore with
seaweed. The TEA average
distance is 120 km each way. This
distance corresponds to a 16,000-ha
farm occupying a square of seafloor
that is 30 km on a side to supply a
CleanCarbon-Energy 8,000-
barrel/day facility.
The transit energy expenditure is
less when the barge harvests and
fertilizes several structures on one
shore-structures-shore trip.
Structural strength is adequate for
maintenance vessels to moor to any
spoke-float or the center float, if draft
TEA calculations in tabs ‘NetEnergy, Mow-HrvOps-Large’ and ‘Mow-HrvOps-Small’ found the energy required to move seaweed
through a 250-m long conveyor and keep the conveyor circling is about the same as the propulsion energy for a small barge during
harvesting. (A small barge is sized to harvest one structure per shore-structure-shore trip. Its draft must be shallow to closely follow
the gang mower. 250-m length for the Large barge conveyor is conservative because the average length would be half that.)
The Large barge operation has not been optimized. There is considerable room for improvement and some issues, including: if the
full barge and tug draft too much, it would have to be moored outside the structure and use a longer conveyor. Alternatively, a portion
of the structure may be low or lower temporarily to allow access to the center.
Table of ContentsBack to 16
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations 4 Mow-harvest-transport Earlier mowing designs
Initial Phase 1 mowing systems
presumed a harvest bag supported in
a catamaran that would follow a
mowing device closely and that cut
Gtik floats quickly (it doesn’t).
Possible cutting mechanisms
included a spinning knife and a band
saw. The band saw is pictured at left
behind rotating guide bars intended
to keep animals away from the saw
and prevent the saw from cutting
The mowing catamaran, pictured at
left, would dewater the Gtik with
conveyor action pushing the
seaweed into an impermeable bag
with a porous end. Consultations
with ASV Global about typical
return-on-capital tow speeds and
the expense of buying bags led us
to favor barges with tugs (or self-
propelled barges).
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations 5 Mow-harvest-transport – Nth scale mowing action
We considered a simple vacuum pump suction into hydraulic conveyance, but did not like using energy to
move all the water that enters the ‘mouth’ behind the trimmer. The snowblower-action is a combination
conveyor and dewatering device. The wastewater and stormwater industry term is ‘thickener’;
screwpresses, centrifuges, vortex inducing shapes. The shaftless screw of a snowblower can be
configured to act as a screwpress. Other mechanisms may work better. The TEA calculations of the
energy required to hydraulically convey the mixture of cuttings and water as much as 250 meters to the
Large stationary barge are based on thickening the cuttings to 2% solids. 2% solids pumps well (almost
as well as water).
After considering other cutting
actions, shear, as used by scissors
and hedge trimers, became the
basis of the final TEA. Hedge
trimmer action is inherently safer
than spinning knives or saws.
Nothing larger than the opening
between blades can be cut.
Openings for Gtik may be less than
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations 6 Planting
Individual Gtik plants can live as long as 40 years (Engel et al. 2001 ). Therefore one planting may be enough for the
life of the structure. The sensitivity analysis uses 5, 10, and 20 years between re-planting. The ‘Planting’ tab
calculates 20x20-cm plant spacing is about right for Gtik plants that will be mowed when about 400 wet grams/plant
(6 wet kg/m2) (Lapointe 1990) .
Shield Gtik from herbivory
Cuttings inside shields and
planted on panels for Phase 2
At Nth scale, the cuttings, or seeded bits of
twine, are planted as part of the pre-
fabrication. The pre-planted panels could
be entire ‘pie’ sections planted in long
narrow seawater tanks. The panel would
scroll from an empty roller to a planted roller
while shielded cuttings are placed between
scrolls. Phase 2 may use rigid ‘frames’ or
net panels for both tank testing and at-sea
trial growing.
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations 7 Mowing, nutrient recycle, on-reef storage seasonal variations
These graphs from Dr. Penniman’s 1983 dissertation
illustrate how much seasonal variability there is in the
nature and content of Gtik grown in New Hampshire. Brian
Lapointe has observed % dry weight varied between 7-10%
depending on availability of nitrogen and growth rate. He
says, “But that 10% seems to be a good number if the
plants are growing and not highly N-limited.”
Observations that may influence operations: The mow-
harvest-transport and the ‘front-end’ of the energy process
handle wet biomass, but oil production is based on dry ash-
free biomass and carbohydrate. Levelized harvesting
(same biomass every day year-round) and on-reef storage
are more important the more the seasonal variation. For
example, the energy process can make more oil from the
same amount of wet biomass harvested in winter than from
that harvested in summer. The energy process would also
have less ash and more inorganic nitrogen leftover from
winter-harvested Gtik.
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Operations 8 Submerging and re-surfacing
16 of 17 spoke/central floats are charged
with air up to 15 atm (will not crush).
All floats have solar panels.
Central float has two (redundant)
compressors and the air release/charging
valves and hose connections for all the
adjustable ballasts and the charged spoke
floats. Simple control. All the valves are in
the central float and controlled remotely or
autonomously by wave height sensors.
Spoke to
central float
Anchor line to
subsurface buoy
4. Surfacing starts when controls
in the central float direct
compressed air from the spoke
floats into the adjustable ballasts.
After a little of the water is
displaced (at 6 to 11 atm abs)
and in the absence of current, the
system rises. Water exits as the
air expands.
Table of ContentsBack to
1. Submerging starts, at far and center left,
by opening air release valves
2. Water rushes into the adjustable ballast as
air leaves or is compressed such that
interior pressure remains near ambient.
3. In the absence of significant current, the
ballast must be nearly full of water before
the system starts to submerge.
The magnitude of waves and currents have significant implications
for the scale and cost of mooring infrastructure. All oceanic regions
are characterized by a unique set of extreme waves and currents:
magnitude, form, orientation, and stratification. In order to
accurately model our design, we first had to analyze the expected
conditions for our intended deployment area in the central Gulf of
Historical records of wave and current measurements were
collected and analyzed from nearby NDBC buoys and the Army
Corps. WIS wave hindcast data points. We also investigated API
standards* for extreme wave and current conditions (for a given
return period and water depth). Since the API recommendations
were most conservative, these estimates were used. The extreme
conditions applied represent conditions present in a Category 5
* American Petroleum Institute: API RP 2SK 2005 “Design and Analysis
of Stationkeeping Systems for Floating Structure” Appendix K
API recommendations pictured above.
On the left, current records on an oceanographic buoy
from Hurricane Andrew show the current shifting
direction dramatically while maintaining large
magnitude throughout the water column.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 1 Oceanographic environmental analysis
Table of ContentsBack to
Above graph data from: USACOE Wave Information Studies
Quasi-static analyses were initially performed to estimate extreme
loads on the initial set of design configurations. These estimates
yielded specifications for structural components. These
specifications were used to define the geometric and material
parameters for initial dynamic modelling scenarios.
Quasi-Static Analysis considers “static” loads from currents and
approximates dynamic wave loads. The quasi-static model
developed provided a platform for evaluating the loads, deformation
and submergence of any given anchor line. It takes into account
chain weight, rope elasticity, float buoyancies, ballast weights, and
current and wave loads as a function of depth.
Because geometric configurations are quick to change, and results
are produced instantaneously, such quasi-static analyses were well
suited for initial analysis of design concepts, sensitivity analyses
and preliminary optimization.
Plan view representation of
loading associated with
lead (upstream) anchor line
The model solves for load equilibrium Graphical Output of the Quasi Static Model
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 2Preliminary Engineering Analysis
Table of ContentsBack to
Using the results of our environmental analysis we determined a set of scenarios that could represent worst case loading.
Because structural loading is highly dependent on the relative magnitudes of ocean waves and currents and specific design
limitations (such as the load limit at which shellfish will be dropped from the system) the worst case loading scenario is not
obvious. We pursued a stepped approach in which each successive set of scenarios revealed which conditions would impart
higher structural loads. In this way we were able to keep the number of modelled scenarios to a minimum.
Since extreme wave and current can come from any
direction during hurricanes, any anchor line has the
potential to be the extreme loaded anchor. Thus,
currents and waves were assumed to propagate in the
same direction (along any arbitrary anchor line)
We chose storm conditions with a 50-year return
period, accepting moderate risk (highlighted yellow) for
a system with a minimum design lifetime of 20 years.
The design loads correspond to a Category 5
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 3Dynamic Modelling Scenario Specification
Table of ContentsBack to
The AdjustaDepth structure was evaluated for its performance and expected loading using the time-domain
finite element analysis software package Hydro-FE, which is capable of fully accounting for geometric and
material nonlinearities. Andrew Drach (Callentis Consulting) constructed and ran these models.
Gracilaria canopy drag load estimates were based on empirically derived estimates for the drag of a rough
plate. The netting cultivation structure and the associated macroalgae were approximated with aggregated
high drag linear elements spanning the spokes, representing the drag of several net panels combined. This
approach is typical for modelling of aquaculture net structures. Hydro-FE estimates hydrodynamic loads on all
structural elements (floats, ropes, ballasts, etc.) using Morrison’s equation.
In order to simulate real world ocean conditions, the structure was subjected to statistical wave sets as
defined by the Jonswap spectrum (scaled with the given significant wave height and period). In order to
ensure consideration of extreme loads, all scenarios were run until load results indicated asymptotic stability.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 4 Dynamic Modelling Results 2
The structure is discretized into
elements with specific geometric,
hydrodynamic and material properties
The structural model with Phase 1 tension management was
subjected to extreme wave and current conditions.
Case 4 Results Cases described in PPT and
3 TEA tabs starting with ‘Env.Dynamics
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 5Dynamic Modelling Results
Table of ContentsBack to
The results showed how the
system remains stable and
relatively unaffected at depth
even in category 5 hurricane
conditions (see profile video).
However, the results also
indicated low tensions on
downstream anchor lines in high
current events. The updated
anchor line design should
improve tension management
and prevent the mooring line
floats from surfacing.
Plan view
Dynamic Modelling results were used to evaluate the
performance of and loading on the structure. We investigated:
Static tensions and geometries
Maximum loads in extreme wave and current events
Substrate motion (sensitivity to passing waves)
Sensitivity to presence of shellfish cultivation
To view videos: Copy and paste google drive web
site address in your browser. Explorer seems easier.
Click on AVI files
Profile view
Extreme load limits on structural
components were determined through
Dynamic Modelling. We then applied safety
factors to these loads as suggested in the
document “Design Criteria for Offshore
Macroalgae Growing Systems (Fredriksson
et al. 2019)”. These recommendations are
informed by the American Petroleum
Institute API-2SK “Design and Analysis of
Stationkeeping Systems for Floating
Structures”, which is ubiquitously used in
the Gulf of Mexico offshore oil and gas
industry. The safety factors used are given
in the table to the right.
Float sizes were chosen according to the
performance (i.e. submergence, stability)
observed in the dynamic models.
All costs were estimated by scaling
material weights (for the given
specification) against cost trends observed
in “off the shelf” items with analogous uses,
scales, and strengths.
Full Scale Structure Component Specifications
Component Safety Factor
Structural ropes, chain, connectors intact 1.67
Structural ropes, chain, connectors damaged 1.25
Drag anchors intact 1.5
Drag anchors damaged 1.0
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 6 Equipment Specification
Table of ContentsBack to
Model Results
Max Anchor line load : 400 kN
Max Spokeline load: 300 kN
Max Anchor load: 400 kN
Max Perimeter line load: 70 kN
Modeling Scenario
25 yr wave and current event
10.3 m sig. wave height
1.45 m/s current
Colinear waves and current
The original 100m x 100m Phase 2 structure with
Phase 1 tension management was modelled
numerically in Phase 1 (shown top right). The results
indicated inadequate performance with the given
specifications (see performance in extreme
conditions bottom right). Nonetheless, the results
provided a reference for the magnitude of loading.
The results were scaled to estimate the equipment
requirements for a smaller phase 2 structure that
would exhibit better performance (with adequate
safety factor). These specifications were used to
estimate the cost of the Phase 2 structure. Early in
Phase 2 we will model the updated phase 2 structure
(with its improved anchor line geometry) to confirm
adequate performance and specifications.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 7 Phase 2 Structure Modelling
Table of ContentsBack to
Profile view of static submerged system
Profile view of submerged system in extreme conditions
Plan view, still photo from video of
submerged system, extreme conditions,
Phase 1 tension management.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Mechanical resiliency confirmation 8 long-life materials
The structure achieves long life using synthetic ropes with several features:
Samson makes long-life rope using creep and fatigue-resistant fibers wrapped with filter tape and covered with
a jacket. Expensive at $100/m, but the filter tape excludes both sediment and biologic sharps while the jacket
protects the filter tape and stops UV. (Biologic sharps are tiny creatures that get into the fibers and then form
shells.) Samson also indicates the tension fatigue resistance of DM-20 turbo rope is superior to fiberglass or
carbon fiber rope.
Applied Fiber terminations (shown at right on simple double braid) provide 100% of rope breaking strength.
In addition to saving money on large installation costs, risks, and interrupted growth; long life ropes are much less
likely to cause loose gear. Suppose we planned for a relatively short life (say 5 years) and a hurricane passed by
in Year 1. Rope exposed to stirred-up sediments could be loaded with sediment sharps. The sharps may degrade
the fibers such that many ropes would break during a hurricane that occurs in Year 5.
Core-dependent double
braid, DM-20 Turbo with
jacket over filter tape
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Equipment use efficiency
Except when using on-structure storage, the flexible floating reef is mow-harvested when the biomass density reaches about 8 DMT/ha (8
wet kg/m2). It is mowed back to 3 DMT/ha (3 wet kg/m2). Thus, if the growth rate is 3.7%/day, each structure would be mow-harvested
every 30 days. Each structure would yield 5 DMT/ha/harvest (750 wet MT). Annual harvest would be 67 DMT/ha/yr. A mow-harvest-
transport (and nutrient return) system moves as one.
If 120 DMT/ha/yr production were level year-round, 5 systems would be needed, not including downtime for maintenance. Equipment
use efficiency for each situation:
Max ambient 45% utilization (5 systems/11 systems) Full seasonal variation of monthly average growth rate from 1%/day (85
days between harvests) in November to 17% (7 days between harvests) in June. Maintenance during the winter.
Semi-Levelized 70% utilization (5 systems/7 systems) Harvest intervals range from 31 days to 12 days. Maintenance during
the winter. The TEA low-cost&energy biomass rate is based on this semi-levelized arrangement, see tabsNetEnergy, Mow-HrvOps-
Large’ and ‘Mow-Harvest$’.
Levelized at winter (60 DMT/ha/yr) rate 80% utilization (2.5 systems/3 systems). Maintenance every six months for each system.
Biomass is stored on the reefs to ensure constant (weather permitting) supply of the energy process. The ‘above level’ primary
production during the summer feeds the fisheries co-products.
Calculation of LARGE barge equipment use efficiency,
-ha farm
Sep Oct
120 DMT/ha/yr
with Max ambient nutrient supply, number of
-harvest-transit systems needed each month
120 DMT/ha/yr
with Lapointe, semi-levelized production,
number of systems needed each month
60 DMT/ha/yr
with Lapointe result levelized
at daily production
equivalent to winter sunlight, but with recycled nutrients
Not likely possible
-Some way to raise wintertime production
such that we could harvest at the rate of
120 DMT/ha/yr
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Deployment area rationale
Rationale for deploying in the Gulf of Mexico:
Large area of open ocean with frequent tropical storms proves economics for even larger areas.
The Mississippi River dead zone provides ample nutrients, making the reefs an environmental benefit.
Gulf oceanography data and research is extensive and relatively fine-grained.
Regulators and citizens around the Gulf are more practical about permits than in most U.S. locations.
Extensive fishing and offshore oil industry equipment and labor expertise can support rapid deployment.
Substantial investment in oil pipelines and refineries that can be kept in business with macroalgae-to-oil.
Phase 2 location
6:1 area ratio in array
Data from NOAA Gulf AquaMapper
Table of ContentsBack to
6 million ha available seafloor between 50 to 100-m seafloor
depth could produce over 100 million DMT per year
AdjustaDepth Phase 1 Final Report DE-AR0000916
Scalability1 General
Phase 1 developed for 1 million ha of substrate on 6 million ha of NOAA-suggested ‘available’ seafloor in the
Gulf of Mexico. Many innovations are generally applicable to 300 million ha of U.S. continental slope.
Seafloor depth
The TEA structure cost is based on 75-
m seafloor depth to provide an average for the 1 million ha of substrate
between 50 to 100-m depth. The area may expand if hurricanes are survived on shallower seafloors or costs
are acceptable on deeper seafloors due to extreme anchor line loads being horizontal.
Tropical storms at the 50 to 100-m seafloor depth in the Gulf of Mexico would appear to have the most
extreme conditions of any U.S. continental shelf system. The current velocity is relatively steady but the
current direction can rotate 180º as the hurricane passes overhead. Computer dynamic testing considered
two extreme events while the structure is submerged Wave-dominated: 17-m significant waves, 28-m
maximum wave, and 1.4 m/s current; and Current-dominated: 12-m significant waves, 20-m maximum wave,
and 1.8 m/s current. This 50-yr event is for the central Gulf. Because the water can be warm, the design rope
is creep-resistant.
Gracilaria tikvahiae (Gtik)is native to the Gulf of Mexico. Gtik is not particularly sensitive to low or high
salinity. Its optimum summertime clear-water growing depth appears to be near 5-m depth. At Nth scale, if
attached and swimming filter feeders do not sufficiently clarify the water, the substrate depth can be shallower.
Gtik has evolved to quickly absorb ammonia to take advantage of large fish swimming by. Up to fist-sized
(invasive) barnacles quickly colonize any structure. The structure is designed as a semi-permanent reef that
will reach a bio-
diverse equilibrium of barnacles, planted shellfish, and volunteer species. The planting system
is designed to maintain growing Gtik after incidents of green sea turtle herbivory.
Infinitely scalable The system design has inexpensive nutrient reuse and distribution from the energy
process for high and/or levelized yields throughout the entire 1 million ha of substrate With only ambient
nutrients, well more than 100,000-ha of substrate between 50 to 100-m seafloor depth is available along the
outer edges of the Mississippi River dead zone. Ambient edge-of-dead-zone nutrients, when combined with
filter feeders converting nutrients from organic to inorganic appear sufficient for average yield above 100
DMT/ha/yr over scales in excess of 100,000-ha. Using dead zone nutrients would make seasonally peaky
yield (high in summer, low in winter), but would have environmental benefits.
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Scalability2 Seafood could pay for structure and monitoring
Seafood price High Low
Scalability constrained by U.S. seafood
imports, conservatively estimated
metric ton/yr
2 3
NOAA attributed statements. See TEA 'ScalingW
Selected value for seafood at the dock,
volume < U.S. imports
$/metric ton
$4,000 $2,700
A half or a third the NOAA valuation. A third is close
to Costa Rica export value.
Necessary income/area of floating flexible
reefs to pay for seaweed
$/ha/yr of
Far above the possible income from seaweed
energy. (based on IMTA and FAD research)
Corresponding reef area with high
seafood productivity needed for profit
tons/ha/yr 38 56
-Production' tab suggests 150 MT/ha/yr is
Maximum reef area before seafood
production exceeds U.S. imports
hectare 53,000 54,000
Gulf of Mexico seafood can pay for the seaweed
operation up to about 50,000 hectare of reef.
Wintertime reef seaweed productivity sets
-round harvest rate to CleanCarbon-
Energy (
CCE) energy processing facility
30 60
There is no reason to harvest more than can be
sustained every day, year
-round, when supplying
less than 3 of the largest CCE facilities.
Income from seaweed, if sold to CCE at
the CCE minimum price
$/ha/yr $6,000 $4,800
3% of the seafood income. The seafood operation
could pay for the entire seaweed operation.
Similar calculation for global seafood
demand, at lower price point
About double the deployment area available in
the Gulf of Mexico with seafood paying half the
Above table is a calculation of the scalability for the concept of ‘seafood pays significant structure and monitoring costs.
Synergies include:
CleanCarbon-Energy (CCE) (like most every energy process) needs the same mass of feedstock every day, 365 days per
year, except when they shut down for maintenance or natural disasters.
Because the seafood product scales substantially at 30 times the income of the seaweed-for-energy, we should design the
harvest system to match wintertime growth rates year-round. Use on-reef seaweed storage to increase levelized harvest.
180,000 m3is one week of storage for CleanCarbon-Energy’s largest facility. 6.3 million m3is 8 months of storage 8 months
of fermenting biomass 50-m high, 200-m wide, and 630-m long, or contained in 1,600 Large harvest barges. (In contrast,
harvesting 50,000-ha at year-round constant rate of 60 DMT/ha/yr requires only 150 tug-barge-mow systems.)
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Nutrients1 Ambient nutrient availability applied to growth model
Dr. James prepared computational growth models for Gtik
growing in the Gulf of Mexico. Data relating growth rate to
nutrient concentrations were sparse as were data on
nitrate and phosphate concentrations in the Gulf of
Mexico. While developing the model, he estimated that
optimum summertime clear-water depth for Gtik was about
5 m. (We started Phase 1 expecting 1 m would maximize
yield.) Dr. James and biologists anticipate seasonal
biologic “triggers.” That is, equivalent conditions in the
spring and fall will not result in equivalent growth.
Moreover, growth rate will depend on the size of the plant
over and above self-shading effects.
Growth rate and yield based on the maximum (blue) model result
Jun Jul
Growth rate maximum
Harvested that month, max (total)
Maximum growth rate (blue curve) is considered conservative because the model only included inorganic nitrate and not ammonium.
Oceanographers and wastewater engineers suggest there is much more organic N than inorganic N nearshore. AdjustaDepth, as a
semi-permanent reef, supports fish and other creatures that convert organic N to inorganic N, thus increasing nitrate and ammonium.
Table of ContentsBack to
AdjustaDepth Phase 1 Final Report DE-AR0000916
Nutrients2 Ambient light intensity
Dr. James’ model includes sunlight intensity, temperature, salinity,
and nitrate & phosphate concentrations. The shapes of the curves
on the graph are thought to result from a combination of
summertime sunlight and spring-summer supply of (excess)
nutrients from the Mississippi River. It does not take into account
possible biologic growth timing triggers due to the lengthening day
in the Spring.
The graph of sunlight intensity with depth on the right indicates very
high turbidity, perhaps microalgae, occurs in June and July,
corresponding to the growth model peaks. While the least turbidity
occurs in October and November, corresponding to the growth
model minimums. We will use this light-penetration information to
set the at-sea depth of the Phase 2 structure and to inform our
monitoring of water clarity as an indication of the effectiveness of
our filter feeders.
Table of ContentsBack to
Latitude: 29.39701446
Longitude: -88.35026079
AdjustaDepth Phase 1 Final Report DE-AR0000916
Nutrients3 Mass-balance check of nutrient shadow
-balance table for Final Report units w-o
Available N concentration in arriving water, all inorganic N, mostly nitrate
µg/L 10
Available N concentration in arriving water, nitrate plus convertible organic N
µg/L 100
Area of substrate growing macroalgae
ha 100,000 100,000
Area of seafloor footprint with AdjustaDepth spacing
ha 600,000 600,000
Length of the side of a square farm
km 77 77
Current that conveys inorganic and organic N into, through, and past the forest array
average current over a year, generally the same direction and perpendicular to a side.
m/sec 0.1 0.1
Thickness (depth) of mixed layer, ideally a nutrient
-weighted average over the whole year meters 30 30
Annual average dry fraction of N contained in Gracilaria tikvahiae (Gt) at time of harvest.
%3% 3%
yield throughout the forest based on mass balance of the N DMT/ha/yr 24 230
The ammonia in the energy process wastewater is equivalent to this many people,
The TEA tab ‘Yield-massBalance’ is a check on the possibility of a nutrient shadow when relying only on ambient inorganic
nutrients. In the simplified table above, a steady current carrying 10 µg/L of nitrate-N into a 100,000-ha of reefs would average 24
DMT/ha/yr. That is far below our expectation for more than 100 DMT/ha/yr and suggests significant nutrient shadow.
AdjustaDepth is designed to carry and accumulate filter feeding finfish and shellfish on a semi-permanent structure. The filter
feeders should provide organic to inorganic nutrient conversions preventing nutrient shadow (at this scale).
Nutrient flux out of the energy process can be a bigger problem than nutrient shadow What to do with the wastewater from the
energy process? 130 million people is 40% of the U.S. population. Wastewater treatment costs range from $50/person/yr to
$600/person/year. (The energy process might pay $280 to $3,400/DMT for wastewater treatment unless nutrients are
None of the issues quantified in the table above affect AdjustaDepth because AdjustaDepth includes inexpensively
recycling the nutrients from the energy process, as described on the next slide.
Table of ContentsBack to
Self-propelled gang
mower and nutrient
distributor spirals
around the structure
Temporarily moored bar ge
Seaweed and nutrient
conveyor tubes
AdjustaDepth Phase 1 Final Report DE-AR0000916
Nutrients4 Detail of recycling nutrients
The transit energy
expenditure is less when the
barge harvests and fertilizes
several structures on one
The nutrient return and distribution happens concurrent with and looks like mow-harvest-transport. Synergies of this particular
combined nutrient recycle and mow-harvest technique include:
With harvest levelized to be the same every day, the interval between harvests will not be less than a month, that may match Gtik
nutrient storage time-capacity.
Each mowing leaves 25% of a full-grown Gtik plant to absorb nutrients needed to triple in mass (mow at 75% of full growth).
Tiny incremental cost on the mow-harvest-transport: a small pump; a hose, and a non-sloshing flexible water tank for nutrient
Need relatively little prep of the energy by-product water, perhaps only anaerobic digestion of the short-chain fatty acids and
Table of ContentsBack to 37
AdjustaDepth Phase 1 Final Report DE-AR0000916
Biomass production cost1
Table of ContentsBack to
$150,000/ha/yr Seafood with
same seaweed harvest every day.
When seafood revenue pays for structure, initially, seaweed cost is as low as $30/DMT up to 3 million DMT/yr
within the Gulf of Mexico, at which point US seafood demand is satisfied. See Scalability2.
$5,000/ha/yr Seaweed for energy,
same harvest every day, 365
days/year for even flow to energy
processor. (Picture area is proportional
to income/area.)
Income per hectare from seafood for
people is about 30 times that of
seaweed for energy
$150,000/ha/yr Seafood with
same seaweed harvest every day.
Later, seaweed cost is less than $130/DMT up to 100
million DMT/yr when scaling to meet global food
demand. See Scalability2 and TEAScalingW-
(Note: Seafood is worth much more to a billion
people with higher incomes. The other 7 billion
people cannot pay as much for food. As food
production increases seaweed-for-energy must cover
more of the structure and monitoring cost.)
AdjustaDepth Phase 1 Final Report DE-AR0000916
Biomass production cost2
Table of ContentsBack to
Issues when considering AdjustaDepth growing Gracilaria tikvahiae in the Gulf of Mexico with the only
income from seaweed delivered to an on-shore energy production facility, based on the smallest
CleanCarbon-Energy facility (450 DMT/every day, 160,000 DMT/yr):
The ‘all middle’ $290/DMT assumes 80 DMT/ha/yr with 10-year structure life. Our growth modeling and
calculations indicate that a wintertime growth rate of 3.7% is possible and would mean harvesting 183
kg/ha/day every day (67 DMT/ha/yr). See TEA ‘Yield-growthParameters’ Rows 90-96. On-reef storage
(live biomass) of summertime production may allow us to harvest more than 220 kg/ha/day every day (80
DMT/ha/yr). That combined with 20-yr structure life predicts $180/DMT. CCE estimates they can afford
to pay $200/DMT for biomass with their smallest unit. (Both CCE’s ability to pay $200/DMT and
AdjustaDepth’s $180/DMT cost do not need seasonal storage.)
We have yet to optimize the substrate cost. The ‘all middle’ number is $1 million per 20-ha structure.
Something as simple as the regulators allowing our use of a 30x200-cm rectangular grid could reduce
capital and perhaps planting costs substantially.
The TEA contains assorted reality checks such as: reductions for mowing efficiency, the days weather
allows harvesting for each month of the year, fraction of daytime spent at optimal depth each month, etc.
We may reduce costs further as we optimize many items such as: the size of the harvest barge.
Without seafood income, biomass production costs should be between $120 to $290/DMT. (These are
‘all low’ and ‘all middle’ costs respectively. Terms explained in Capital slide.)
AdjustaDepth Phase 1 Final Report DE-AR0000916
Biomass production cost3
Table of ContentsBack to
The terms: ‘all low’, ‘all middle’, and ‘all high’ are explained in the Capital slide and appropriate for understanding the first two
‘Biomass production cost’ slides, the Critical assumptions, and the Sensitivity analysis slides.
The table above provides a quick overview of how the projected biomass changes from best case to worst case with and without
seafood revenue paying for some portion of the structure, monitoring costs, and nutrient return/distribution. In all cases the seaweed
operation would pay for planting Gtik in shields and the full cost of mow-harvest-transport. If the seafood were to pay 100% of the
structure costs, the biomass-for-energy cost would be $30/DMT.
The Sensitivity analysis starts with the ‘all low’ situation and then changes the appropriate assumption in the TEA. This means that it
does not show the $1,000/DMT “all high” situation.
Installed capital costs and total costs for all three scenarios
1,000-ha aggregate perennial aquaforest
units all low all middle all high
Structure: anchors, ropes, floats, ballasts, connections, netting,
and installation
$$110,000,000 $140,000,000 $190,000,000
Mechanical (depth adjust), electrical (solarPV), monitoring, and
control on the structures
$$6,700,000 $7,600,000 $8,900,000
-harvest-transport units , rounded (See 'Mow-Harvest$'
Row 46 for variations)
$$38,000,000 $36,000,000 $40,000,000
Tot al initial capital, rounded
$$150,000,000 $180,000,000 $240,000,000
Total cost per t on, delivered to shore
-seaweed pays all $/DMT $120 $290 $1,000
Fraction of seafood payment for structure
%75% 50% 25%
Total cost per t on
seafood, at 30 times the value/area pays
for structure, monitoring, and nutrient return. Seaweed
pays for planting and mow
-harvest, transport.
$/DMT $50 $180 $760
AdjustaDepth Phase 1 Final Report DE-AR0000916
Energy Return Energy sensitivity and critical assumptions
Table of ContentsBack to
Energy harvested per MHT trip
Using 2,433 heat of combustion for Gtik
Energy to mow and convey to barge,
mow of 4 structures
Power calculations in kW, converted to energy in
diesel using 30% engine
-generator electrical
efficiency, can be improved to 40% with waste
Energy for round trip transit
Energy for maintenance trips
Estimated as fraction of MHT energy
Energy recovery ratio
One return
Solar PV to air compressors
Above energy return calculation is abbreviated to allow room for a tornado plot specific to energy return. The parameters in
the tornado plot also represent the critical assumptions associated with the energy return ratio. The complete calculation is
in the TEANetEnergy, Mow-HrvOps-Large’ with the condensed table on Rows 102 11 2.
Note MHT is Mow-Harvest-Transport energy
Bold indicates number used for EER=38
Average travel distance from farm to CCE
energy processing facility: 15, 115, and 200 km
Pumped solids vary from 0.5, 1, 2%
Maintenance energy 5%, 10%, 20%
MHT system is either Large stationary barge
or Small following barge.
10 30 50 70 90 110 130
MHT system (large-small)
Maintenance energy
Pum ped % soli ds
Distance to CCE
ERR Sensitivity Analysis, XX:1 ratio
High or bold ERR
Middle or bold ERR
AdjustaDepth Phase 1 Final Report DE-AR0000916
Critical assumptions 1
Table of ContentsBack to
biology and Ecology
Assumed % all solids wet,
water shook off
10% 10%
% solids affects estimating the optimum wet density before and after mowing but not
the dry yield estimates. 1976 LaPointe study reports 10% solids for Gtik
Biomass density, just
before mow
3.0 2.5
Refer to tabs JamesWkst and Yield
growthParameters for details of calculated yield.
The pre
-and post-mow densities are selected to yield 4 wet kg/m2
with each mowing
based on Lapointe 1978, 1990 and Capo
-Lapointe 1999 experiments with Gtik
indicating an optimum range of plant density for the greatest yield. All three
supported high sustained yields, with the 1999 4
-year experiment finding a mean of
DMT/ha/yr (converted from g dry wt/m2/day). Dr. Scott James' growth model
suggests min, average, and max %growth/day for each month of the year. The
growth rate is based on insufficient data combining inorganic and organic nitrogen
within a few 100 kilometers of the proposed 1,000
ha aggregate farm. Although three
lab experiments suggest a sustained high of 145 DMT/ha/yr is achievable with
recycled nutrients, the 'low
cost' 120 DMT/ha/yr matches the limit when only ambient
nutrients are used in the growth model.
Biomass density, just after
8.0 7.5
Calculated yield, rounded
(Actually the result of
research and modeling
outlined in 'Yield
ha /yr
80 50
Time between replanting
Gracilaria tikvahiae (
20 10 5
Literature based on a few years' genomic analysis of life stage suggests individual
plants live and grow for 40 years. The real questions
-Do mowed plants keep
growing upward to accommodate more mowing? (Big YES for self
-planting kelp
forests. Don't know for
Gtik.) Animals or disease might eat plants such they won't
Estimated fauna co
production when not
harvesting macroalgae
150 100
See TEA ‘CoP
-Production’. Suppose there is sufficient N and P to grow 100
DMT/ha/yr (~1,000 wet tons/ha/yr) of seaweed. The fauna consists of more
concentrated N & P. Loss is mitigated by the shear number of co
products spanning
the food chain from sea cucumbers to apex predators. (
Scalability analysis
40 wet tons/ha/yr might earn $150,000/ha/yr in the U.S. Gulf of Mexico.)
Fraction of flexible reef
capital cost paid by seafood
50% 25%
See logic under the nutrient recycling figure in "Summary, w
-CoP" tab.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Critical assumptions 2
Table of ContentsBack to
NOAA scalability analysis for GoM
-100 m seafloor
MM ha
5.8 5.8
The NOAA graphic showing a ribbon area available for flexible reefs
in the Gulf of Mexico averaging about 80 nautical miles (150
from shore. Square arrays sized to feed the largest (8,000
barrel/day output)
facility would have a seafloor
footprint about 30 kilometers on a side. One side of the square
array would be on the 50
-m depth contour. The harvest barge trips
would be straight
-line hypotenuse to points in the square. See
calculations in TEA tab 'CleanCarbon
Distance to energy facility, one way
(Distances are calculated
on placing a CleanCarbon
(CCE) facility anywhere along the
U.S. Gulf Coast.)
km 62 115 168
Transit speed over ground to and
from energy facility
m/s 2.6 3.1 3.6
Speed estimates are based on phone conversations with ASV
Global. We asked for their quick thoughts on optimum transit speed
for an autonomous barge
-tug considering both energy and capital.
(Results of this conversation is one of the reasons we abandoned
harvesting into bags as something to model in the TEA at this time.)
5.0 6.0 7.0
generator efficiency (fuel to
Charts of fuel consumption per kWh are readily available online from
manufacturers. Larger engine
generators are more efficient than are
smaller. Full load is more efficient than is partial load.
Density of
cuttings dropped into
the barge
The density of
Gtik dropped into the barge is based on wastewater
engineering experience for piles of drained biomass.
Energy Return Ratio sensitivity CCE could install their facilities off shore in the center of each square array of reefs.
Doing so would boost our energy return ratio above 100:1, not including embedded energy. However, CCE currently
estimates the additional cost (and cost is often proportional to embedded energy) of working from offshore platforms is too
much more than an on-shore installation.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Critical assumptions 3
Table of ContentsBack to
Economic life of grow
substrate and framing rope
10 10 5
Service life is varied in the '
' tab, 20, 10, and 5 years. It appears one has
two choices: either the reef is "disposable" with perhaps a 3
-year life or it should and can
be built for a 20
-year economic life. This because synthetic rope without filter tape and
jacket won't be able to survive the design event after about 3
5 years. Rope degrades due
to UV, creep, fatigue, and sharps cutting fibers. (Other rope materials, such as fiberglass
and carbon fiber, may have less issue with sharps but more issue with fatigue.) The
sharps can be either sand/silt/clay or biologic. Tiny critters getting into the fibers and
growing sharp shells can only be prevented with filter tape. Filter tape has to be covered
with a jacket. Discussions with Dr. Krueger
Hadfield, June 2018, confirmed we can count
on individual plants living for 10 to 40 years through many mowings. That implies
connecting and disconnecting the main ropes from the grow
-substrate and/or 'framing
ropes' once a decade or less.
Economic life of large rope
& connections: mooring,
spoke, and circumferential
(aka catenary shellfish)
20 15 10
Economic life of buoys and
20 15 10
Economic life of anchors
and chain
20 20 20
Maintenance expense on
the structure expressed as a
% of the capital investment
per year
% of
That is: A structure costing $100 million would spend 1%/yr or $1 million/yr on
maintenance. Managers of properties often budget 1% of the property value per
year for maintenance. https://
-hour days between
structural down
40 30 20
This is the minimum time between storms (large enough to submerge). The higher the
number of days, the smaller is the compressor and the solar PV capacity and the lower
their capital cost. If we need to submerge without full compressed air tanks, a service boat
can supply compressed air after the storm. With typical current, a tiny burst of air into the
downstream ballasts will cause them to surface. If no current, then very little compressed
air surfaces the whole structure. After the ballasts rise, the compressor can top
off with air
using solar power.
-hour days to push out
remaining ballast water
when near surface
Nominal installed power
based cost rate for U.S. Gulf
of Mexico coastal areas
Installed costs for Solar PV are decreasing. The adjustment factor is Mark
Capron's professional engineer thought for the additional installation cost for
streamlined mounting of panels on the surface floats with a clear super hydrophobic
Adjustment factor to power
cost rate for robust streamlined
mounting on floats
2 2 2
$50 $100 $150 $2 00 $250 $300
Seafood pays some
MHT cost vs Barge Size
MHT Yie ld leveling
Fuel cost
Structure maintenance %
Re-seeding interval
Structure service life
Substrate cost
Yield per area per time
Cost ($/DMT of macroalgae transported to on-shore energy process)
Sensitivity Analysis, $/DMT
AdjustaDepth Phase 1 Final Report DE-AR0000916
Sensitivity analysis1 Variations explained in next slide
Table of ContentsBack to
Details in Biomass production cost3
and Sensitivity analysis2 tabs
Low cost
Middle cost
High cost
Yield per area per time
Yields DMT/ha/yr (120, 80, 50) affect on $/DMT.
Substrate cost
$ ($500k, $1m, $2m/structure) different estimated costs for the combination of framing
rope with grow
substrate netting. This is difficult to estimate for small macroalgae, like
Gtik, because the optimum density requires close parallel rope spacing or a grid near
-cm square. Regulators are signaling a tighter grid, smaller than 10x10-cm
square. On the other hand, a small grid enables automated (as opposed to
autonomous) mow
This chart starts from the ‘all low’ situation for $/DMT. The TEASensitivityTornado’ tab includes one parameter
(barge size & system) for energy return ratio as well as pie charts.
AdjustaDepth Phase 1 Final Report DE-AR0000916
Sensitivity analysis2 3 ranges of key variables in previous slide
Structure service life
Service life, Years (20, 15, 10) affect on $/DMT. Need the filter tape and jacket for 3+year
life due to sediment stirring from an early storm and possible storm at end of service life.
Mechanical resiliency slide.
seeding interval
Years (20, 10 , 5) between re
-seeding, effect on $/DMT. Gtik plants are long-lived.
maintenance %
Maintenance expense at 1, 2, or 3% of structure capital per year. (1% is $1 million/yr on
$100 million capital.)
Fuel cost
Cost of marine bio
-diesel: $4, $6, $8/gallon.
Yield leveling
harvest Transport (MHT): Range of growth %/day (0%, 6.4%, 16%) effect on $/DMT.
(The difference in growth rate between peak summer month and low winter month.) Only
considering the LARGE barge. Nutrient recycling supports some yield leveling for reduced
-harvest capital costs. See discussion of yield leveling and TEA ‘NetEnergy, Mow-
MHT cost vs Barge
-harvest Transport: Barge size (4-structures/trip, 1 structure/trip) effect on $/DMT.
Larger barges save some on capital and fuel cost. But barge size is limited by draft.
Seafood pays some
Fraction (75%, 50%, 25%) paid by non
-seaweed effect on $/DMT. If or how seafood
production might drop off when seaweed is harvested such that its density changes
between 3 to 8 wet kg/m
2, is unknown. The biology might work out to keep seaweed
production steady (say mowing every 25 days, year round) while non
-seaweed products
grow more during the summer such that seaweed
-for-energy pays only for its mow-
transport costs. More discussion in the TEA ‘Summary, w
-CoP, Rows 9-14.
Table of ContentsBack to
Process Description:
Hydrothermal liquefaction (HTL) of biomass within a well completion to provide
high pressure, temperature, and cross-flow heat exchange
At >374ºC and 25-MPag, water at supercritical conditions acts as an effective
solvent/partial hydrogen donor to generate biocrude and biochar.
At True Vertical Depth (TVD) 2000-3200 m hydrostatic head provides pressure
increase to supercritical conditions localized at reaction zone.
Exceptionally long cross-flow heat exchanger within a concentric completion
provides required process efficiency.
On-shore and offshore applications.
Biocrude (approx. 40% yield) of a low sulfur bunker fuel valued at a premium to
West Texas Intermediate oil ~$600/MT Marine Gas Oil (highest class of marine oil.)
Biochar (approx. 60% yield) can be sold as a blending feedstock for existing
power plants ~$240/MT.
Key Advantages:
Ability to process wet biomass and recycle nutrients
Reduced steel wall thickness for low capital cost
Based on existing oil and gas technology, suited for U.S. Gulf Coast
Small surface facility footprint
Lower operating cost with effective heat exchange
Containment of reaction zone
AdjustaDepth Phase 1 Final Report DE-AR0000916
Path-to-Market CleanCarbonEnergy1 Downhole Supercritical Water Liquefaction (SCWL)
On-Land Deployment
SCWL can also float in
up to 3 km deep water
HTL in a well
Table of ContentsBack to
Slide provided by CleanCarbon.Energy
450 DMT/Day
On-shore New Orleans
Pilot Demonstration
>10,000 DMT/Day
Four CCE on-land
1,000,000 Barrels of
oil per Day (BOPD)
AdjustaDepth USGC
AdjustaDepth Phase 1 Final Report DE-AR0000916
Path-to-Market CleanCarbonEnergy2 Scaling on U.S. Gulf Coast
CAPEX: $20-30 million
Location: New Orleans, LA (or equivalent) on Air Products hydrogen pipeline
Feedstock: Barged in municipal waste, wood biomass, available macroalgae
Objectives: Demonstrate Downhole SCWL of ‘waste’ and macroalgae, demonstrate
nutrient recycling to Ocean Forests, achieve cash flow for R&D
Locations: Initially near Gulf Coast hydrogen pipeline
Feedstock: Ocean Forest produced seaweed biomass at 3% of
the value of Ocean Forest produced seafood
Objective: Commercial operation while demonstrating integrated
robust supply to SCWL of macroalgae and nutrient recycling for
complete reef-to-reef manufacturing
Location: U.S. Gulf Coast (USGC)
Feedstock: Ocean Forest Macroalgae
Objective: Full commercial deployment paying less
than $100/DMT for macroalgae feedstock
Table of ContentsBack to
Slide provided by CleanCarbon.Energy
AdjustaDepth Phase 1 Final Report DE-AR0000916
Path-to-Market CleanCarbonEnergy3 first 20,000 barrels of oil /day
Initial Commercial Deployment:
Reinvest profits from CCE demonstration facility using ‘waste’ and available seaweed as feedstock
Reinvest profits to expand Ocean Forests’ seafood and seaweed-feedstock operations
Capital Intensity $20-30,000 CAPEX/BOPD; comparable to Canadian Oil Sands Projects.
Barge wet biomass to wells, recycle nutrients from SCWL process backhauled on the same barges
Increase Ocean Forests’ biomass harvests ahead of SCWL facility completions
Potential hydrogen plant additions
Market other products: electricity; biochar; fertilizer; asphalt; hydrogen facility optimizations
Table of ContentsBack to
Slide provided by CleanCarbon.Energy
AdjustaDepth Phase 1 Final Report DE-AR0000916
Path-to-Market Ocean Forests4 Developing country infrastructure
Ocean Forests are seeking humanitarian funding to support UN Sustainable
Development Goals in developing countries, because of the expected high seafood
yields with large profits to provide food and jobs, while cleaning up ocean pollution,
which reduces disease and increases tourism. An analysis of how Ocean Forests
directly address 12 of the 17 UN Sustainable Development Goals is posted here.
Spreading Ocean Forests to developing countries serves ARPA-E goals when major
components are manufactured in the U.S.A. (because U.S. products have long
service life) and when U.S. foreign aid supports local infrastructure that reduces the
need for people to migrate to the U.S.
Examples of Countries Currently Interested
in Installing Ocean Forests
Costa Rica
Dominican Republic
Tan za nia
Table of ContentsBack to
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.