A Sea Mixer is a large nuclear-powered ship that is outfitted with an even number of symmetrically-aligned, very-large, downward-directed sub-surface mixer propellers.
The picture shown here is the CAD construct used in our computational fluid dynamics (CFD) simulations, the change in color at the base of the hull shows the position of the waterline in our simulations – the mixing blades are positioned 40 m below the water surface, and are coupled to the ship’s steam turbines by drive shafts and gearing compatible with the required propeller power and rotational rates (in our simulation ~50 MW per mixer shaft).
The ultimate objective is to drive an enormous plume (or jet) of warm surface waters to depths of 150-400 m as efficiently as possible. (“Jet” would be the normal term, but that doesn’t seem quite right for a multi-stream flow at only 2-3 m/s.) There are two keys to achieving this: (1) The mean plume velocity a short distance below the propellers should not be much more than needed (a few m/s) to balance the buoyancy head of a 300-400-m column of surface water in the sea’s mean density over that depth, and (2) the plume’s mean diameter at 100-m depth should preferably be ~100 m. Achieving this requires total downward propeller shaft power of 100-800 MW, depending primarily on the ship’s cruising speed, the Mixing Layer Depth (MLD), and the desired sea penetration depth.
Greater Detail
Figure 1 is a front (bow) external view of our four-propeller Sea Mixer model. In it one immediately sees the two large downward-directed front mixer propellers 101, 102 along with a bow view of a rough CFD CAD stand-in for a Nimitz-class aircraft carrier.
In the upper stern-port-direction perspective view of Figure 2 the two downward-directed rear mixer propellers 203, 204 can be seen along with some other features. Again, in order to produce a downward plume of very high mass-flow rate as efficiently as possible, the flow velocity leaving the propellers should be no more than needed to balance buoyancy and viscous factors, the latter of which are extremely low and the former of which are rather low at least to the depth of the MLD. To that end, the plume’s mean initial flow velocity need only be 2-3 m/s. To achieve meaningful flow rates, the mixer propellers should have diameter greater than 50 m.
For reference, we included in our simulation model the four propulsion propellers on Nimitz-class carriers. Two of these propulsion propellers 205, 206 can also be seen in Figure 2, but they are more clearly visible in a later figure.
Figure 3 is a Sea Mixer drive-train schematic drawing combined with a stern-side (rear) external view of the model ship. It illustrates how two steam turbines 301, 302, driven by a nuclear reactor 330, could be coupled to the two rear mixer propellers 203, 204 by clutches 303, 304, high-speed drive shafts 305, 306, speed-reduction gear boxes 307, 308, and high-torque vertical drive shafts 311, 312. The drive trains could be much more complex than shown, including additional turns, drive shafts, support bearings, and speed reduction gears. Nimitz-class aircraft carriers have two nuclear power plants, and the additional components of the nuclear power plant are not shown.
The vertical drive shafts would be supported at least by upper bearings 321, 322 and lower bearings 323, 324. The bearings would be secured to the ship by suitable support structure, such as seen in support beams 331, 332, 333, 334, 335, 336. The locations at which the support beams are connected to the ship’s hull 340 will likely need to be strengthened by additional beams along its surface.
(Obviously, for practical reasons both the model and the schematic were made to be as simple as possible while still presenting a reasonably complete representation).
For coupling of the required power to the propellers, the high-speed drive shafts would operate at rotational rate typically 20 to 200 times higher than that of the propellers. The required high-torque gear boxes may push gear technology somewhat beyond prior industrial experience. However, that would essentially be a matter of larger planetary and bevel gears than seen in the nacelles of the largest wind turbines, where torques can exceed 2E7 Nm, which is two orders of magnitude beyond the torques seen in the axles of the largest tractors and an order of magnitude beyond the total torques seen in the wheels of large locomotive engines.
To permit the operation of the sea mixer propellers at practical torques, the mixer propellers would require low relative pitch compared to what is typically seen in ship propulsion propellers. (Mean pitch is normally defined as that at 75% of the propeller radius. Relative pitch is defined as the ratio of mean pitch to propeller diameter.) Propulsion propellers typically have relative pitch in the range of 0.5 to 1.5, depending on the expected typical cruising speed and wake fraction of the vessel as well as motor power and gearing design.
The mixer propellers should have a relative pitch in the range of 0.1 to 0.5 to permit operation at powers in the 25-150 MW range with torques that are not more than several times those seen in the largest currently available wind turbines. Hence, they would operate at higher relative tip speed (the ratio of tip speed to the mean stream velocity leaving the propeller) than is seen in ship propellers. Maximum tip speed of the mixer propellers will likely be ~ 40 m/s, and the propellers would likely operate in the 6-14 rpm range.
Sea Mixers would not be able to enter most ports or drydocks with their mixer propellers mounted. The mixer propellers, some portion of the vertical drive shafts, and likely the lower bearings and some portion of the lower support structure, would need to be mounted in a special sea-mixer service platform offshore in deeper waters after leaving drydock. Before returning to drydock or a normal port these would need to be dismounted at an offshore service platform, which could be a rather complex operation not to be undertaken any more often than necessary. But as a nuclear-powered vessel, fuel and waste transfer would be on the order of only a few tens of tons every few years and could easily be managed at sea. Personnel, maintenance equipment, food and disposable goods transfers could likewise easily be managed at sea – transferred ship-to-ship or via helicopter. The mixer propellers would likely have lower blade solidity (defined as the ratio of blade area to total swept area) than what is typically seen in propulsion propellers – primarily to reduce mass so that mounting and dismounting are more manageable. (The lower solidity also helps improve efficiency in propellers operating at high relative tip speed.)
Of course, the mixer propellers could be damaged from encountering an unexpected obstacle. The Sea Mixer would then need to return to an offshore service platform for propeller replacement.
It should be noted that the Sea Mixers would employ multiple methods to minimize encounters with whales, dolphins, sharks, and large sea turtles. Scout vessels and drones would be deployed to assist in advance detection and diversion of large sea animals. Sonar could be used to detect such, and both lights and sounds could be employed by the Sea Mixer and support vessels to possibly cause them to swim away. The sea mixer ship could decelerate and otherwise adjust its course if necessary. If a collision with a large or endangered animal appeared imminent, the sea mixer could quickly power down the mixer propellers.
The Sea Mixers need to have a landing deck at least sufficient to accommodate multiple large helicopters so that it may not need to re-enter port for many years. The use of a repurposed aircraft carrier accommodates this quite easily, but future potential Sea Mixer designs would have to take that into consideration and leave sufficient room amidst whatever additional marine biology, oceanography, meteorology, and climate related research labs and equipment gets piggybacked onto each deployed Sea Mixer in the gradually growing fleet.
From the combination of lower relative pitch and lower solidity than generally seen in ship propellers, the mixer propellers may operate with somewhat lower propeller efficiency than typical of propulsion propellers. Considerable further blade optimization will come from CFD simulations in concert with system optimization, and the end result may look somewhat like a cross between those seen on trolling boats and those found in quadcopter drones.
When the mixer propellers power up (assuming four propellers, ~70-m diameter and ~40-MW power per propeller), the load on the ship from the propellers, their drives, and their support structures would change from perhaps ~12,000 tons load to a net load of only ~2000 tons, and the ship would rise – causing ~9% reduction in its displacement. If a mixer propeller then fails during operation (from encountering an obstacle, or from some drive or component failure) the load on the ship would become asymmetric and cause the ship to slowly begin to roll. (Carriers, because of their width, have enormous moment of inertia about the longitudinal axis as well as the other axes). To stop this slow roll before it possibly becomes an unacceptable list (tilt), the ship’s stability control systems would quickly power-down all the turbines, possibly by reversing the direction of the steam flows to them, as is sometimes done in these ships for rapid deceleration or quick maneuvers. (Steam is supplied separately from the reactors to additional steam turbines in these ships for electrical power generation to keep it independent of propulsion.)
The clutches in the mixer propeller drive trains could be designed as controllable slip clutches that limit peak torques in the drive train components in the highly unlikely event of an abrupt impact to prevent or limit damage to various components (including the steam turbines) and to safely dissipate large amounts of stored energy (in the angular momentum of the various components) when necessary. The stability control system could adjust steam as needed to functioning propellers to quickly return the ship to normal attitude and stability.
During normal operation, the ship’s stability system would be able to rapidly adjust power on each mixer blade to maintain stability even in rough seas (similar to how a quad-copter drone can maintain stability in moderate winds). The ability to control the power independently to the various mixer propellers gives the ship the stability of a trimaran, should the need ever arise in high seas. While the torques and forces from the mixer propellers would be high and the downward currents all around the ship would be rather high, they would be steady. With suitable control, the stability of the 110,000 t mixer ship should be similar to what is typical for aircraft carriers in relatively calm seas. (The controls could even be balanced so as to correct the 7 degree list that Nimitz class aircraft carriers have when they are deployed – due to the asymmetric mass and area of the flight deck and control tower).
While the sea-mixer ship would never navigate shallow waters even close to the depth of the propellers and would be equipped with the sonar and radar, a scout ship would still always go ahead of the mixer ship to insure there was no obstacle in the path of the ship (such as a sunken ship standing on end in relatively shallow waters). Additional layers of protective measures would be included: Shear pins and hinges at several locations along the lengths of the mixer propeller blades could limit damage from some impacts. Shear pins in the connections of the support structures to the bearings and to the ship could prevent major structural damage to the ship in the highly improbable event of a direct encounter of a mixer hub with a massive obstacle, such as an abandoned oil rig.
Future Sea Mixers – designed and produced for this exclusive mission rather than repurposed from retiring aircraft carriers – might have more pairs of downward directed mixer propellers. (The number of mixer propellers should be even so that that they could balance their respective torques and keep the net torque on the ship to a minimum).
A smaller size or smaller propeller diameter could still be effective, it would merely require a slower cruising speed and a more narrow mixing path. For instance: With only 40 MW per mixer propeller, the initial plume velocity may not be sufficient to send the plume much below ~200 m depth at a reasonable cruising speed if the MLD is only 20 m because of the high buoyancy head from the denser waters below the MLD. That, however, is not a problem. With each successive pass the MLD steadily deepens because the deeper waters are steadily warmed from the mixing of previous passes, steadily reducing the buoyancy head and allowing the plume to penetrate deeper, which then mixes and warms the deeper water in the nearby regions that the Sea Mixer will traverse in subsequent passes.
All four mixer propellers 101, 102, 203, 204 are seen in the bottom external view of our model Sea Mixer illustrated in Figure 4. The four propulsion propellers 205, 206, 407, 408 are shown approximately as they are positioned on an aircraft carrier. The typical waterline 409 is seen around the portion of the hull normally submerged, which is shown in darker shading. The flight deck 103 viewed from the bottom is shown in light shading, and an intermediate shading is applied to the portion of the hull above the waterline. Longitudinal beams 428, 429 between the adjacent lower bearings assist in providing the needed support stiffness.
The arrows near the mixer propellers in Figure 4 indicate rotational directions of the propellers, which here have alternate clockwise/counterclockwise directions so that the net torque on the ship from driving the mixer propellers is very small and easily within the normal azimuth control ability of the ship by its propulsion propellers and rudder.
Figure 5 shows flow vectors (of lengths proportional to velocity, in light blue color) and a color plot of the z-component of the velocity field from a steady-state computational fluid dynamics (CFD) simulation of our model shown in the previous figures. The results shown are from a yz plane a short distance in front of the two rear mixer propellers, in which the coordinate system has the x direction along the length of the ship and the z direction is perpendicular to the sea surface. The propellers here are 80-m diameter and of a simple non-optimized four-blade design, 40-m below the surface, operating at 10.8 rpm in sea water of nearly uniform temperature. The bottom of the plot shown is at a depth of ~420 m, and Vz there is still negative.
The CFD results show that four large low-pitch propellers can send a 2-3 m/s plume of surface waters down to 400 m after the MLD is nearly that deep, which confirms the validity of the concept: A practical amount of ship power (100-600 MW) can drive the needed mass-flow magnitude so that a small number of nuclear sea-mixer ships can essentially end cyclone genesis in the Atlantic. Preliminary propeller calculations indicated that with optimized 70-m propeller blades, 50 MW per propeller would produce a downward plume of over 40,000 t/s. That is about 10 times the flowrate over Niagara Falls (but at a MUCH lower velocity).
When the mixer propellers are operating there would be strong inward directed currents (~1 m/s) all around the ship toward its hull, where they turn downward and accelerate. Hence, for safety reasons, it will be important to have a secure fence to at least shoulder height surrounding all exposed decks, as rescue of any person that might go overboard into the sea (for example, if on deck during a storm) would be improbable. It might be desirable to have a safety net surrounding the carrier as an extra and obvious safety measure, if only to reduce stress levels in crew members that might be on deck while the mixer propellers are operating.
Repurposing the Nimitz Class Aircraft Carriers
There are 10 Nimitz-class carriers in the US fleet, and every 3 to 4 years, another is scheduled to be decommissioned (at a cost of about $1B) and replaced by a new Ford-class carrier with 260 MW of nuclear power and more advanced warfare and communications features. It may take a fleet of six to ten retro-fitted and repurposed carriers to keep SST below 25°C in all regions of high cyclone genesis index in the Atlantic. Moreover, there is strong incentive to put an end to Atlantic hurricanes as quickly as possible. Toward that end, higher shaft power is desired. The Ford-class carriers currently in production achieve 25% higher power primarily from improved optimization of the steam turbines, the condensers, and other associated equipment. The same changes could be made in repurposed Nimitz-class carriers giving them a substantial increase in their shaft power even without changing out their reactors, though that too might be warranted because of their age.
In addition to adding the mixer propellers and their associated drive shafts, gears, support structures, and control systems; there will also need to be substantial changes in the drive trains to the propulsion propellers. It would be necessary to have independent control over the power going to the propulsion propellers while the mixer turbines maintain continuous maximum available power. The easiest way to accomplish this would be to disengage the steam turbines from the propulsion propellers and drive them with electric motors.
The 64 MWe of electrical power available in Nimitz-class carriers is sufficient, as it should require only ~20 MW to propel the ship at 25 km/hr, and ~40 MW to propel the ship at ~30 km/hr. If the power plant (steam turbines, condensers, etc.) of a repurposed Nimitz-class carrier is being upgraded, its electrical generating capacity could also be upgraded to something approaching the 102 MWe of Ford-class carriers. This would allow 20 MW motors on each of the propulsion propellers, which would allow the Nimitz-class carrier to cruise at up to ~50 km/hr while still applying more than half of its power to the mixer propellers.
Tropical depressions will continue to form until SST or mid-troposphere relative humidity have been sufficiently reduced over a very large percentage of the regions of high cyclone genesis index. By maintaining the option for high ship propulsion power, when a depression is beginning to form, the nearest sea-mixer ships could move quickly to begin forming cool patches ahead of its likely path. By executing optimum mixer paths ahead of the developing storm it might be possible to cool a portion of the surface waters feeding the storm sufficiently to shut it down. Whether or not this would often be successful in the first few years of operation remains to be seen, as the radial inflow winds feeding the developing storm often extend for more than 500 km and the mixer ships can do little to affect SST over such a large area in a short period of time.
The drag on the mixer propellers when cruising at high speeds becomes substantial, leading to high forces on the bearings supporting the vertical drive shafts, but such forces would still be about two orders of magnitude below the dynamic load limit of available spherical roller bearings for example, which go beyond 3E7 N. In the figures the supports to the lower bearings are shown slightly above the bottom of the ship’s hull so that they do not add to the ship’s draft when the mixer propellers and lower portions of the vertical drive shafts are removed. They could be positioned lower to reduce stresses on the bearings, but that would complicate compatibility with drydocks and is not necessary.
As Ford-class carriers will continue to be in production for decades, a logical approach would be to accelerate the planned replacement rate of Nimitz-class carriers (modernize the carrier fleet more quickly) and outfit these carriers with larger reactors and more efficient turbines in addition to adding the other needed features for deep sea mixing. The cost of doing so is not likely to be significantly more than the expected decommissioning cost, which then could be deferred for many decades.