Can Nuclear Subs Excel Over Diesel Electric Boats

 

Naval Institute Style · Technical Note on Propulsion Acoustics

PROCEEDINGS

Vol. LXXXII Annapolis · Independent Forum No. VI

Acoustic Signature · Reactor Architecture · Generation IV

The Coolant Pump Problem

Battery-electric propulsion is physically quieter than any pressurized-water reactor because the PWR cannot stop circulating primary coolant and the battery has nothing moving at all. The question for the next twenty years is whether molten salt reactors will rewrite that inequality.

By Stephen “Pseudo Publius” · Third in the German-Diesel Series · Technical Note

—— BLUF ——

A well-built battery-and-electric-motor propulsion train has no continuous primary machinery noise at patrol speeds. It is, by construction, quieter than any pressurized-water reactor, which must maintain forced circulation of its primary coolant under normal operating conditions. This is a physical constraint of the PWR fuel-rod architecture, not a design limitation that can be engineered away.

Natural-circulation PWRs (U.S. S5G, S8G, S9G; Russian OK-650 and KTP-6) partially defeat the problem by eliminating coolant-pump operation at reduced power settings, using thermosiphon convection instead. U.S. Ohio- and Virginia-class and Russian Yasen-class boats exploit this for quiet-transit tactics. But the forced-circulation regime returns at higher power, and the reduction gearing and steam-cycle machinery never go quiet.¹

Molten salt reactors (MSRs) operate at atmospheric pressure, use natural circulation inherently at much higher coolant densities than water, and can be engineered to eliminate primary coolant pumps altogether. A 250 MWt thorium-fueled MSR submarine propulsion design was published in 2024;² the U.S. Molten Chloride Reactor Experiment (MCRE) at Idaho National Laboratory produced its first fuel salt in December 2025, with operational testing planned for 2030.³ The acoustic implication is substantial: if MSR submarine propulsion matures, the “AIP diesel is quieter than nuclear” axiom may cease to hold in the 2040s.

The first two articles in this series treated the diesel engine as the defining component of the conventional submarine. This note treats the coolant pump as the defining component of the nuclear submarine. In a PWR, the pumps that circulate pressurized water through the primary loop are the single largest source of continuous, detectable machinery noise. In an Ohio-class SSBN they run the width of the reactor compartment. They cannot be turned off at mission-relevant power. They are the reason an AIP diesel-electric boat on batteries is, in the littoral acoustic envelope, harder to find than a nuclear attack submarine.

I · The Architecture of Silence What the Battery Does Not Do

Consider the propulsion train of a submerged, battery-powered diesel-electric boat operating on patrol. The battery banks, typically lead-acid in older hulls and lithium-ion in newer Japanese and Korean boats, have no moving parts. Current flows from the battery bus through solid-state power conversion to an electric motor. That motor turns through reduction gearing, or in modern boats through a direct-drive low-speed double-armature configuration, and drives the propeller or pump-jet.

Moving parts in this train: the motor armature, the reduction gear (if present), the shaft bearings, the propeller. Sources of continuous mechanical noise: the shaft bearings and the propeller, both of which are also present in nuclear submarines and are not differentially relevant. The motor itself, properly built, is acoustically near-silent at the power levels used for quiet-transit operations (two to four knots). There is no combustion. There is no steam cycle. There is no primary coolant. There is, most importantly, no pump that must run to keep the boat from cooking itself.⁴

The only caveat is that “well built” is doing significant work in that description. Submarine battery banks produce heat under discharge and require hotel ventilation. Cooling pumps for onboard air-conditioning and sonar electronics run continuously. Hydraulic pumps serve the steering and diving planes. All of these are vibrational sources. The modern German, Japanese, and Swedish approach is to mount them on acoustic rafts with elastomeric isolation, treat the bulkheads with Alberich-type anechoic tiling, and run the hotel bus at reduced load during patrol. The boat is not silent in an absolute sense. It is quieter than the ambient noise floor of the ocean at frequencies where sonar listens.⁵

II · Why the PWR Cannot Do This The Primary Coolant Constraint

A pressurized-water reactor cannot stop circulating its primary coolant during power operations, because the fuel-rod architecture relies on forced convection to remove the fission heat from the cladding surface. If primary circulation stops at full power, the fuel rods overheat in seconds. The pumps that circulate primary coolant are therefore mission-critical, and their operation is continuous. In a U.S. S6W (Seawolf) or S9G (Virginia) reactor, two large primary coolant pumps operate throughout normal power operations, each producing shaft-rate tones, blade-rate tones, and broadband cavitation and flow noise that propagates through the primary loop and into the hull.⁶

The U.S. Navy has fought this problem for sixty years. The S5G reactor plant developed for USS Narwhal (SSN-671), commissioned in 1969, introduced natural-circulation primary cooling: the core is placed low in the primary loop, the steam generator is placed high, and the density difference between hot rising water and cold descending water drives circulation by thermosiphon without requiring pumps at low-to-moderate power. The S8G reactor for the Ohio-class SSBNs, designed by General Electric and fielded from 1981, took this further and can operate at a “significant fraction” of full power on natural circulation alone—reportedly above 70 percent in some sources.¹ The S9G in the Virginia class retains the natural-circulation capability.⁷

This is a genuine acoustic advance. An Ohio-class SSBN on a deterrent patrol, running at four to six knots in natural-circulation mode with the primary pumps secured, is about as quiet as a pressurized-water reactor can be made. The Russian Yasen-class SSGNs, with their fourth-generation KTP-6 reactor built by Afrikantov OKBM, use the same approach and have been described by U.S. Naval Forces Europe Commander Admiral James Foggo as “very quiet, which is the most important thing in submarine warfare.”⁸ Russian submarine doctrine assumes the KTP-6 and the earlier OK-650 achieve full-power natural circulation; U.S. sources are more conservative.⁹

But natural-circulation PWRs do not solve the problem; they mitigate it. At sprint speeds above fifteen knots or so, the primary pumps must engage. Once they engage, the reactor plant's acoustic signature climbs to a level that is detectable at useful ranges, and the boat reveals itself in exchange for the speed. A nuclear boat is quiet when it is loitering and loud when it is closing. A battery-electric boat has the same trade-off (quiet on batteries, loud on diesels) but with a different tactical geometry: the diesel runs only when charging, and charging is done at snorkel depth, out of contact.

Acoustic Profile by Propulsion ArchitectureContinuous machinery noise at patrol speed (2–6 kt submerged)

Architecture	Primary Loop	Continuous Pumps	Tactical Note
Battery-electric (SSK)	None	Hotel only	Quietest mode available; battery-limited endurance
AIP (Stirling, fuel cell)	Low-ΔT closed loop	Minor auxiliary	Near-silent; 2–3 weeks submerged
Forced-circulation PWR	High-pressure water	2 primary + aux	Continuous pump signature at all speeds
Natural-circulation PWR	High-pressure water	Aux only at low pwr	Pumps engage above ~15 kt; S8G, S9G, KTP-6
MSR (proposed)	Atmospheric-pressure salt	None (natural circ.)	Theoretical; MCRE experiment 2030

III · The MSR PropositionWhat Changes With Molten Salt

A molten salt reactor operates under a fundamentally different thermodynamic regime than a PWR. The primary coolant—typically a fluoride salt mixture such as LiF-BeF₂ (FLiBe) or a chloride salt for fast-spectrum variants—is liquid at atmospheric pressure across a temperature range of roughly 450°C to 1,400°C. Operating pressure is approximately 1 bar, compared to 150–160 bar for a PWR primary loop. The density difference between hot and cold salt is larger per degree than for water, which means thermosiphon-driven natural circulation is structurally more robust.¹⁰

Three acoustic consequences follow directly.

First, primary coolant pumps become optional by design. The MSRE (Molten Salt Reactor Experiment) at Oak Ridge in the 1960s used pumps, but subsequent design work—including Oak Ridge's own 1970s liquid-fluoride thorium reactor (LFTR) proposals and the 2024 Indonesian-authored 250 MWt submarine MSR concept published in the Journal of Applied Research and Technology—has demonstrated pump-free configurations in which the primary loop circulates by natural convection at all power levels.² The 250 MWt submarine concept uses a LiF-BeF₂-ThF₄-²³³UF₄ fuel salt in a virtual one-and-a-half-fluid configuration, beryllium oxide moderation, and a high-density graphite reflector, with calculated iso-breeding at breeding ratio 1.057 and temperature coefficient of reactivity of −2.56 pcm/K.²

Second, there is no high-pressure primary loop. The loud, flow-induced noise produced when pressurized water passes through the primary loop restrictions, valves, and heat exchanger headers does not have a direct analog in an atmospheric-pressure salt loop. Flow velocities can be significantly lower for the same thermal power throughput, because the volumetric heat capacity per unit mass of molten salt is comparable to water, and the density is higher.¹¹

Third, the steam generator and steam-cycle machinery on the secondary side still exist if the reactor drives a Rankine cycle. This is the constraint that must be honestly stated: MSRs eliminate the primary pumps but not the secondary turbines, reduction gearing, condenser seawater pumps, or feedwater pumps. If an MSR-powered submarine uses conventional steam propulsion, the acoustic improvement is real but limited to the primary side. The full benefit is realized only with a direct thermal-electric conversion architecture—thermoelectrics, organic Rankine cycles, or supercritical CO₂ loops with magnetic bearings—that eliminates the conventional steam-turbine signature entirely. This is the direction the most recent maritime MSR research is heading.¹²

—— HISTORICAL NOTE ——

The U.S. Navy has tried non-PWR reactor propulsion before. USS Seawolf (SSN-575) was commissioned in 1957 with an S2G sodium-cooled beryllium-moderated reactor (SIR Mark A). The boat operated for two years before superheater-tube leaks and the difficulties of maintaining liquid sodium above its 97.8°C melting point during maintenance led Admiral Rickover to have the reactor replaced with the standard S3G PWR in 1960. The Soviet Alfa-class SSNs used lead-bismuth eutectic-cooled fast reactors (BM-40A / OK-550), producing extraordinary speed and a very small reactor compartment, but requiring continuous shore-side salt-heating infrastructure that neither the Severodvinsk nor the Polyarny yards could reliably provide. All seven Alfas were decommissioned by 1996. The lesson of both programs: non-PWR naval reactors are thermodynamically superior and operationally demanding.

IV · The State of the Art in 2026 Where MSR Maritime Is Today

Four programs of note are active as of this writing.

CORE POWER, the UK-based maritime nuclear firm led by Mikal Bøe, is partnered with TerraPower, Southern Company, and Orano in the Molten Chloride Reactor Experiment (MCRE) hosted at Idaho National Laboratory's Laboratory for Operation and Testing in the United States (LOTUS) test bed. Idaho National Laboratory announced on 4 December 2025 that it had successfully produced the first batch of enriched uranium chloride fuel salt—18 kilograms per batch at 90 percent conversion efficiency—for the MCRE. Operations are planned to begin at INL “as soon as 2030.”³ CORE POWER's “Liberty” program, launched in February 2025, targets floating nuclear power plants (FNPPs) first and civil marine propulsion second, specifically positioning U.S. shipyard production to “compete head-on with China.”¹³

Seaborg Technologies of Denmark has ABS Feasibility Statement approval for a compact Generation-IV molten fluoride salt reactor with a twelve-year refueling cycle, partnered with Korean shipyards for power-barge serial production.¹⁴

China Nuclear Energy Technology Corporation (CNNC) achieved first criticality of the TMSR-LF1 thorium molten-salt reactor—a 2 MWt experimental unit in Wuwei, Gansu—in October 2023. This is the world's only operating MSR as of 2026. The Chinese program is structured to build a 10 MWt demonstration reactor next and a 100 MWe commercial prototype by the 2030s. Naval application has not been publicly stated but is widely assumed to be part of the long-term planning.¹⁵

The U.S. DOE / ABS maritime reactor working group, under the American Bureau of Shipping's $800,000 research contract funded through the U.S. Industry Opportunities for Advanced Nuclear Technology Development program, is evaluating MSRs, heat-pipe reactors, and high-temperature gas-cooled reactors for commercial maritime propulsion. The Herbert Engineering Corp. study commissioned by ABS in 2023 examined nuclear propulsion impacts on a 14,000-TEU container vessel and a 157,000-DWT Suezmax tanker.¹⁶ None of this work is explicitly naval, but the engineering overlap with naval propulsion is direct and the technical lessons will transfer.

Corrosion remains the dominant engineering obstacle. The Oak Ridge MSR Experiment in the 1960s operated its Hastelloy-N primary circuit safely for only a few years at 650°C. Chromium dissolution by fluoride salts above 700°C limits PWR-era nickel-alloy piping. Modern material candidates include advanced Hastelloy variants, silicon carbide composites, and 316H stainless with tellurium-resistant surface treatments; none has yet demonstrated forty-year service life in an irradiated, salt-wetted naval environment. The Mean Time Between Failures record required to field an MSR in a U.S. or Royal Navy submarine under Naval Reactors oversight does not yet exist.¹⁰

V · Proceedings Assessment - What It Means for the Submarine of 2045

The battery-versus-reactor acoustic comparison is real, current, and not about to be repealed in the near term. Modern AIP-equipped diesel-electric boats remain the quietest submarine platforms in the world's oceans. The Götland's JTFEX performance in 2005 was not a fluke; it was the first public demonstration of an asymmetry that has grown more pronounced as Japanese, Korean, and Scandinavian yards have moved to lithium-ion battery banks with energy densities three to four times those of lead-acid, enabling longer quiet-mode patrols at lower thermal overhead.⁵

The natural-circulation PWR is a partial response. The S8G in the Ohio class and the S9G in the Virginia class both exploit thermosiphon cooling at low power, and the forthcoming S1B reactor for the Columbia-class SSBNs will continue the practice. But they do not eliminate the coolant-pump constraint; they work around it at reduced power. A Virginia on natural circulation at four knots is approximately as quiet as the reactor plant allows. A lithium-ion Type 218SG or Taigei on batteries at four knots has no reactor plant to be quiet about.

“Pumps are one of the loudest components in a reactor plant. Spinning impellers, motors, bearings—they all make sound that can be detected by enemy sonar. A natural circulation reactor can operate at low to moderate power without running pumps at all.” — Nuclear Submarine Reactor Technical Guide, October 2025

The MSR proposition is serious but not near-term. The MCRE at Idaho National Laboratory does not operate until 2030 at the earliest. Scaling from a MCRE-class test bed to a 200 MWt naval-propulsion reactor under Naval Reactors certification standards is a fifteen-to-twenty-year engineering effort by historical analogy (S1W prototype 1953 → Nautilus at sea 1955 was extraordinarily fast; the S5G program took eight years; the S8G took nine). A realistic operational MSR-powered submarine in any navy is a 2045-2055 artifact. The Chinese TMSR-LF1 program, moving at a pace U.S. policy analysts have repeatedly underestimated, could compress this; it is not the base case.

The longer-term strategic point is that the acoustic advantage of the conventional AIP submarine is not forever. If MSR naval propulsion matures—if CORE POWER and TerraPower and INL deliver the MCRE on schedule and the thermal-to-electric direct conversion side of the architecture matures in parallel—a 2050-era U.S. Virginia-class successor could plausibly operate with atmospheric-pressure primary cooling, no primary pumps, and a direct-drive electric propulsor. Such a boat would combine the endurance and sprint speed of a nuclear platform with the patrol-quiet acoustic signature of a battery boat. Nothing currently in the world's fleet matches that specification. Nothing on any drawing board in 2026 is less than twenty years from it.

The MSR pathway also, incidentally, reduces the proliferation problem associated with naval PWRs. The Chinese TMSR-LF1 is thorium-fueled, producing ²³³U in-situ with much lower weapons utility than highly-enriched uranium PWR fuel. U.S. naval reactors currently use weapons-grade HEU; the UK Astute- and Dreadnought-class reactors do as well. An MSR navy would not need that fuel cycle. For a United States that has struggled with its HEU supply—the TRIGA/ML-1/BEAR program's repeated difficulties obtaining fresh fuel for submarine refueling—the logistics relief would be considerable.¹⁷

In the meantime, the operational lesson of 2026 remains the one this series has been developing across three articles: the conventional submarine, powered by a German diesel, is not the poor relation of the nuclear boat. In its intended operational envelope, it is frequently the superior platform. The United States has noticed this twice in Proceedings in the last decade and has yet to act on it. If an MSR-based nuclear propulsion architecture does mature in the 2040s, the institutional question will not be whether the United States can acquire it; it will be whether the United States still has an operational submarine force to apply it to, and a domestic shipbuilding base capable of building the hull around it. Neither question currently has a comfortable answer.

◊ ◊ ◊

This is the third of a series on propulsion architecture in the modern submarine force. The first treats the Fairbanks-Morse opposed-piston diesel; the second the Type 209 and German MTU diesel export control; this third treats the coolant-pump acoustic constraint that separates battery-electric from pressurized-water nuclear and the Gen-IV reactor architectures that may eventually close the gap.

Verified Sources & Citations

Primary technical references, peer-reviewed proposals, and federal program documentation.

1. “S8G reactor,” Wikipedia (updated 2026); GlobalSecurity.org technical summary. Documents natural-circulation capability at a “significant fraction” of full power without reactor coolant pumps.
   en.wikipedia.org/wiki/S8G_reactor · globalsecurity.org/military/systems/ship/systems/s8g.htm
2. Hasibuan, M. H., Dwijayanto, R. A. P., Meireni, M., Harto, A. W., Novianto, F. F., & Widiastuti, P., “Core neutronic design of small modular molten salt reactor for submarine propulsion,” Journal of Applied Research and Technology 22(3) (2024), 403–409. DOI: 10.22201/icat.24486736e.2024.22.3.2336.
   jart.icat.unam.mx/index.php/jart/article/view/2336
3. Idaho National Laboratory, “Idaho lab produces first-ever fuel for fast molten salt reactor experiment, opening door to maritime commercial reactor deployment,” 4 December 2025. Molten Chloride Reactor Experiment (MCRE) partnership: Southern Company, TerraPower, CORE POWER, U.S. DOE.
   inl.gov/feature-story/idaho-lab-produces-first-ever-fuel-for-fast-molten-salt-reactor-experiment
4. “Why are diesel-electric submarines quieter than nuclear submarines?” Naval Post technical analysis (2021); battery-electric acoustic profile at patrol speeds.
   navalpost.com/nuclear-submarines-diesel-electric-submarines-noise-level/
5. Brent M. Eastwood, “An AIP Submarine Quieter Than Ambient Ocean Noise 'Sunk' A U.S. Navy Aircraft Carrier,” 19FortyFive, 11 January 2026. JTFEX 2005 HSwMS Götland exercise results.
   19fortyfive.com/2026/01/an-aip-submarine-quieter-than-ambient-ocean-noise-sunk-a-u-s-navy-aircraft-carrier/
6. “Nuclear Submarine Reactor: Complete Technical Guide 2025,” DefenceWing, 9 October 2025. Technical discussion of PWR coolant pump acoustics and natural-circulation mitigation.
   defencewing.com/nuclear-submarine-reactor-complete-guide/
7. World Nuclear Association, “Nuclear-Powered Ships,” accessed 2026. Confirms S9G natural-circulation capability on Virginia-class SSNs and 33-year service-life core.
   world-nuclear.org/information-library/non-power-nuclear-applications/transport/nuclear-powered-ships
8. Peter Coates (“Pete”), “Russian Yasen-class SSGN,” Submarine & Nuclear Matters blog, December 2023. Incorporates public comments by Adm. James Foggo, then Commander, U.S. Naval Forces Europe, on KTP-6 reactor natural-circulation acoustic performance.
   gentleseas.blogspot.com/2023/12/russian-yasen-class-project-885-ssgn.html
9. World Nuclear Association, “Nuclear-Powered Ships,” op. cit. Details on VM-5 / OK-650 / OK-659B / KTP-6 reactor lineage and the fifth-generation supercritical-water reactor (SCWR) reportedly under development for Russian fifth-generation SSNs.
10. “Molten-salt reactor,” Wikipedia (updated 2026). Thermodynamic and materials-science summary, including Hastelloy-N chromium-dissolution limits above 700°C and the OrNL MSRE operational history.
    en.wikipedia.org/wiki/Molten-salt_reactor
11. “Molten Salt Reactors: Military Applications Behind the Energy Promises,” POWER Magazine, 3 December 2018. Thermodynamic comparison of MSR versus PWR primary-loop conditions (1 bar vs. 150 bar; 750°C vs. 300°C).
    powermag.com/molten-salt-reactors-military-applications-behind-the-energy-promises/
12. BMT Defence Services Ltd., “The potential of the molten salt reactor for warship propulsion,” International Naval Engineering Conference (INEC), May 2012. Proposes NaK-cooled drain tank with organic Rankine-cycle auxiliary generators for post-shutdown power continuity; archived by the Thorium Energy Alliance.
    thoriumenergyalliance.com/wp-content/uploads/2020/02/bmtdsl-molten-salt-reactor-confpaper-inec-may12.pdf
13. CORE POWER / American Nuclear Society, “Core Power launches U.S.-anchored maritime civil nuclear program,” ANS Nuclear Newswire, 20 February 2025. Liberty program overview and CEO Mikal Bøe's stated objective of U.S.-yard-built MSR maritime propulsion.
    ans.org/news/2025-02-20/article-6774/core-power-launches-usanchored-maritime-civil-nuclear-program/
14. “Molten Salt Reactors: Maritime's Nuclear Option,” Marine Link, 23 September 2022. Seaborg Technologies / Korean shipyard partnership; ABS Feasibility Statement (2020).
    marinelink.com/news/molten-salt-reactors-maritimes-nuclear-499681
15. “Developments Of Nuclear Energy In Shipping Now,” Ship Nerd News, May 2025. Confirms China's TMSR-LF1 as the world's only operating molten salt reactor as of 2023–2026 and its role as the flagship of Chinese Generation-IV civil nuclear development.
    shipnerdnews.com/developments-of-nuclear-energy-in-shipping-now/
16. World Nuclear News, “ABS awarded federal contract for marine nuclear propulsion project,” 2022; Seatrade Maritime News, “Shipping can't afford to ignore vast potential of nuclear propulsion: ABS,” July 2024. Herbert Engineering Corp. study parameters.
    world-nuclear-news.org/Articles/ABS-awarded-federal-contract-for-marine-nuclear-pr
17. Congressional Research Service reporting on U.S. HEU supply for naval reactors, and the debate over whether a transition to low-enriched uranium (LEU) naval fuel would be operationally viable. Not MSR-specific but directly relevant to the Gen-IV alternative.
18. Supplementary references: BMT Defence Services naval MSR concept paper (2012); MIT / University of Texas DOE research partnerships on maritime advanced reactors; Lloyd's Register and Maritime Executive coverage of modular MSR maritime propulsion 2021–2025.

— PRODUCED IN THE STYLE OF U.S. NAVAL INSTITUTE PROCEEDINGS —
THIRD IN THE GERMAN-DIESEL SERIES · TECHNICAL NOTE · APRIL 2026