MIT Taps High-Field Tokamaks to Beat ITER to Net Fusion Power
MIT Plasma Science and Fusion Center
High-temperature superconducting magnets enable tenfold increases in tokamak power density, allowing compact reactors to achieve net energy gain within 15 years.
Fusion energy offers a zero-emission, high-density, dispatchable baseload power source that avoids the runaway meltdown risks and long-lived high-level waste of nuclear fission. (2:08)
The deuterium-tritium (D-T) fuel cycle remains the only near-term viable option for fusion energy due to its low ignition temperature (~15 keV) and high reaction cross-section. (14:07)
High-temperature superconducting (HTS) tapes, specifically rare-earth barium copper oxide (REBCO), enable on-axis magnetic fields of 9 to 12+ Tesla, allowing tokamaks to achieve net-energy gain at a fraction of ITER's size. (48:49)
MIT's proposed SPARC and ARC designs leverage this HTS technology to target net electricity on the grid within 10 to 15 years, bypassing the multi-decade timeline of traditional international roadmaps. (54:07)
Detailed notes
## Background
This talk was published on YouTube by the MIT Plasma Science and Fusion Center on February 6, 2017. Delivered by Assistant Professor Zach Hartwig as part of the PSFC's Independent Activities Period (IAP) lecture series, it outlines the core physics of magnetic confinement fusion and introduces MIT's proposed pathway to accelerate fusion energy commercialization.
## TL;DR
Hartwig establishes three foundational physics-based rules for assessing any proposed fusion reactor, focusing on the insurmountable hurdles of non-D-T fuels and the empirical dominance of the Lawson triple product. He systematically reviews alternative fusion concepts—from open magnetic mirrors and self-organized plasmoids to stellarators—demonstrating why only the tokamak has approached energy breakeven (Q ≈ 0.65). Rather than waiting decades for gargantuan conventional facilities like ITER to reach net power, Hartwig presents MIT's strategy: deploying commercially mature REBCO high-temperature superconductors to double magnetic field strength on-axis. Because volumetric power density scales with the fourth power of the magnetic field (B^4), this breakthrough enables compact, lower-cost devices like SPARC and ARC to demonstrate net energy (Q > 1) on a timescale relevant to combating climate change.
## Atmosphere & style
The presentation carries the brisk, razor-sharp clarity of an elite academic seminar, combining rigorous physics critiques with grounded, dry humor. Hartwig speaks with an accessible yet unapologetically exacting tone, dismantling speculative alternative concepts and pathological science with whiteboard-clear analogies rather than academic hand-waving. It feels less like an investor pitch and more like a definitive foundational briefing, infusing deep scientific skepticism with urgent, tangible optimism about overcoming fusion's decades-long perception problem.
## Bulletin
- [02:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=128s) Fusion energy offers a zero-emission, high-density, dispatchable baseload power source that avoids the runaway meltdown risks and long-lived high-level waste of nuclear fission.
- [14:07](https://www.youtube.com/watch?v=L0KuAx1COEk&t=847s) The deuterium-tritium (D-T) fuel cycle remains the only near-term viable option for fusion energy due to its low ignition temperature (~15 keV) and high reaction cross-section.
- [48:49](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2929s) High-temperature superconducting (HTS) tapes, specifically rare-earth barium copper oxide (REBCO), enable on-axis magnetic fields of 9 to 12+ Tesla, allowing tokamaks to achieve net-energy gain at a fraction of ITER's size.
- [54:07](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3247s) MIT's proposed SPARC and ARC designs leverage this HTS technology to target net electricity on the grid within 10 to 15 years, bypassing the multi-decade timeline of traditional international roadmaps.
## Key takeaways
- [02:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=128s) Per-capita energy consumption is strongly correlated with human development metrics, but historically this has required burning fossil fuels and emitting carbon dioxide.
- [02:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=128s) Fusion energy provides a unique value proposition: zero greenhouse gas emissions, high volumetric power density, dispatchable baseload capability, no runaway meltdown risks, and no long-lived high-level waste.
- [17:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1074s) Evaluating fusion concepts requires understanding that fuel choice fundamentally dictates the difficulty of achieving net power, with deuterium-tritium (D-T) being the most feasible.
- [18:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1119s) The Lawson criterion establishes that a sustained burning plasma requires a sufficient product of plasma density, temperature, and energy confinement time.
- [24:28](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1468s) Cold fusion (LENR) has no known physical confinement mechanism, was experimentally refuted by institutions like MIT, and is classified as pathological science.
- [32:28](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1948s) Alternative magnetic confinement concepts like magnetic mirrors, field-reversed configurations (FRCs), and spheromaks remain orders of magnitude below the performance required for energy relevance.
- [37:24](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2244s) Stellarators achieve helical magnetic confinement entirely via complex external coils, making them the first non-tokamak configuration to show meaningful progress in the Lawson parameter space.
- [38:17](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2297s) Tokamaks have historically dominated fusion research, achieving energy gain factors up to Q ≈ 0.65 and representing the only configuration near the breakeven threshold.
- [43:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2619s) Tokamak performance progress plateaued in the mid-1990s not due to plasma physics limits, but because conventional low-temperature superconductor technology forced machines to scale to massive, slow-to-build dimensions.
- [46:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2799s) Fusion volumetric power density scales with the fourth power of the magnetic field (B^4), meaning higher magnetic fields allow for much smaller, cheaper, and faster-to-develop reactors.
- [48:49](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2929s) The commercialization of REBCO high-temperature superconductors (HTS) as flexible, high-strength tapes enables on-axis magnetic fields of 9 to over 12 Tesla.
- [51:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3114s) The ARC conceptual design utilizes HTS to target 500 MW of fusion power at approximately one-tenth the volume of the ITER facility.
- [51:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3114s) SPARC is designed as a compact, university- or startup-scale HTS tokamak aimed at demonstrating net fusion energy gain (Q > 1) to break public and institutional cynicism.
- [01:03:21](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3801s) While tokamaks are inherently pulsed due to transformer action, experiments at Alcator C-Mod have demonstrated that radiofrequency (RF) waves can non-inductively drive plasma current indefinitely.
- [01:07:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=4028s) The ultimate engineering limits to increasing magnetic fields in tokamaks are the mechanical yield strength of structural materials supporting immense Lorentz forces and the minimum shielding thickness required for 14.1 MeV neutrons.
## Detailed walkthrough
### [00:00](https://www.youtube.com/watch?v=L0KuAx1COEk&t=0s) Welcome and PSFC IAP Series Overview
The session opens with welcoming remarks introducing the first in the 2017 Independent Activities Period (IAP) lecture series hosted by the MIT Plasma Science and Fusion Center (PSFC). Attendees are invited to a tour of the laboratory facility at 175 Albany Street immediately following the lecture, convening in the NW17 conference room. A slate of upcoming IAP talks is highlighted, spanning diverse facets of fusion and related science: a Friday session featuring two talks and a tour covering the High Energy Density Plasma Physics program at MIT and the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory; an exploration of alternative funding models for fusion research presented by a professor from the MIT Sloan School of Management; and a presentation by Nuno Loureiro on plasma physics and self-magnetization, a phenomenon ubiquitous across the universe. The host also previews an upcoming talk by Harvard's David Keith on climate engineering and geoengineering, remarking with dry humor that “it reads a little bit like something from a James Bond villain, I guess, but he tells me that it's all benign.”
The host then introduces the main speaker, Zach Hartwig. Hartwig earned his Ph.D. in the Department of Nuclear Science and Engineering (NSE) at MIT in 2014, continued as a postdoctoral researcher, and was recently appointed as an Assistant Professor in NSE. His lecture serves as an introduction to fusion energy fundamentals while detailing the PSFC’s strategic vision for magnetic confinement fusion.
[02:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=128s) The Energy-Climate Dilemma and the Value Proposition of Fusion
Taking the stage—and joking about trying not to fall off the elevated platform—Zach Hartwig begins by acknowledging PSFC colleagues Dan Brunner, Bob Mumgaard, and Brandon Sorbom for their collaboration in preparing the presentation, as well as NSE alumnus Todd Rider for contributing material on evaluating disparate fusion approaches. Hartwig anchors the motivation for fusion in the global imperative of energy production. Plotting energy consumption per capita against societal indicators such as child mortality, the UN Human Development Index (HDI), or average income reveals an unmistakable positive correlation. However, Hartwig stresses that plotting those same human development metrics against carbon dioxide emissions per capita yields virtually identical curves, highlighting the central tragedy of modern industrial history.
```mermaid
flowchart TD
accTitle: Comparison of Energy Sources Against Societal Requirements
accDescr: Flowchart contrasting fossil fuels, fission, and renewables with the operational advantages offered by fusion energy.
subgraph Challenges["Conventional Energy Limitations"]
Fossil["Fossil Fuels<br/>CO2, SOx/NOx, Particulates"]
Renewable["Renewables<br/>Intermittent, Low Power Density, Grid Storage Needs"]
Fission["Fission<br/>Meltdown Risk, Long-Lived Waste, Proliferation"]
end
subgraph Solution["Magnetic Fusion Energy"]
Fusion["Fusion Power<br/>High Power Density, Baseload / Dispatchable, Inexhaustible Fuel"]
end
Fossil -.->|"Replaced by clean baseload"| Fusion
Renewable -.->|"Complemented without massive storage"| Fusion
Fission -.->|"Replaced without meltdowns or deep waste"| Fusion
```
Hartwig confronts this reality directly: “There is no other way to do it presently than to dump carbon dioxide into the atmosphere by burning fossil fuels.” Every nation that has achieved high standards of living has done so at the expense of the planetary climate. Fusion energy offers a fundamentally distinct alternative: fusing two light nuclei to liberate net energy, such as extracting deuterium from seawater to generate orders of magnitude more energy than is consumed. When weighed against competing energy sources, fusion exhibits transformative advantages:
- **Versus Fossil Fuels:** Produces zero carbon dioxide, sulfur oxides (SOx), nitrogen oxides (NOx), or toxic particulate emissions; utilizes an essentially inexhaustible, universally accessible fuel supply requiring no environmentally destructive extraction or global shipping logistics.
- **Versus Variable Renewables:** Possesses high volumetric and material power density with a compact land footprint; operates on demand as dispatchable baseload power sited near demand centers; feeds into existing electricity grids without requiring utility-scale storage technologies.
- **Versus Nuclear Fission:** Poses zero physical possibility of a runaway meltdown; creates no long-lived, high-level nuclear waste requiring deep geologic repository disposal; carries practically no proliferation hazard, as it requires neither uranium nor plutonium, and illicit production of fissile isotopes within a fusion environment is easily detectable.
[05:10](https://www.youtube.com/watch?v=L0KuAx1COEk&t=310s) Sorting Winners from Losers: The 1905 Aviation Analogy
Given these immense societal and financial stakes, the fusion landscape has attracted substantial private capital and dozens of distinct technological concepts, each vying to deliver commercial fusion power. Yet the sheer diversity of reactor schemes makes it notoriously difficult for investors, the public, and even working fusion scientists to distinguish between viable contenders, long shots, and unworkable designs. To ground this dilemma, Hartwig presents an analogy from early aviation circa 1905.
```mermaid
flowchart LR
accTitle: The 1905 Aviation Metaphor for Fusion Concepts
accDescr: Flowchart illustrating the distinction between addressing an unsolved physics problem versus solving an engineering problem.
subgraph SchemeA["Concept A: Superficial Engineering"]
A1["Premature Deluxe Features<br/>Engines, fast speed, cabin luxury"] --> A2["Fatal Flaw:<br/>Missing the wing / unproven physics"]
end
Hartwig contrasts two hypothetical investment pitches from that era: one entrepreneur pitches a passenger craft boasting plush cabins and flight attendants serving drinks at 40,000 feet, yet when inspected, the design lacks wings. Hartwig points out the immediate common-sense reaction: “You can't fly without a wing.” By contrast, another team presents fundamental equations of aerodynamic lift and wing shear alongside experimental balsa gliders pushed off hillsides without casualties. The latter group has tackled the foundational physics, leaving the subsequent scale-up as an engineering challenge. Hartwig explains: “The key difference here, you should think about this is a physics problem.”
The goal of the lecture is to equip the audience with three simple physics-based rules to objectively evaluate any fusion concept, followed by an overview of the PSFC's proposed pathway: employing high-magnetic-field tokamaks to achieve net fusion energy.
[08:04](https://www.youtube.com/watch?v=L0KuAx1COEk&t=484s) Nuclear Physics Foundations: Binding Energy
Hartwig initiates the technical framework by posing the first evaluation question: what are the viable fusion fuels, and how do their fundamental properties dictate the confinement approach? Nuclei consist of positively charged protons and neutral neutrons bound together by the strong nuclear force, which acts across minute distances to overcome the mutual Coulomb repulsion of the protons.
Nuclear stability is governed by binding energy per nucleon—the energy required to hold a given nucleus together. Certain nuclei, such as iron, occupy the most tightly bound state at the minimum of the nuclear potential energy curve and require very little additional relative binding adjustment. In contrast, very light elements (like hydrogen isotopes) and very heavy elements (like uranium) have lower binding energy per nucleon. Rearranging nucleons from loosely bound configurations into more tightly bound configurations frees binding energy that can be harvested as useful heat and electricity, either through the fission of heavy elements or the fusion of light ones.
[09:13](https://www.youtube.com/watch?v=L0KuAx1COEk&t=553s) The Physics of Fusion and Evaluating Potential Fuels
Fission and fusion represent two opposite sides of nuclear binding energy. Whereas fission splits massive, heavy nuclei into smaller product nuclei that require less binding energy to hold together—releasing the surplus energy to do useful work—fusion takes light, weakly bound nuclei and pushes them together to liberate binding energy. To evaluate which nuclei can practically be fused for energy, two fundamental physical quantities must be assessed: reaction energetics (the balance of input energy versus output energy) and reaction probability.
```mermaid
flowchart LR
accTitle: Fusion Fuel Feasibility Pipeline
accDescr: Diagram showing how nuclides are filtered based on Coulomb barrier, probability, and energy yield to select optimal candidates.
subgraph S1["Initial Pool"]
A["3,000 Known Nuclides"]
end
subgraph S2["Physics Filters"]
B["Coulomb Barrier Filter<br/>(Exclude Z > Lithium)"]
C["Cross Section Filter<br/>(Exclude Low Probability)"]
end
subgraph S3["Viable Fuels"]
D["Marginal Fuels<br/>(e.g., p-B11)"]
E["Optimal Fuel:<br/>Deuterium-Tritium (D-T)"]
end
A --> B
B --> C
C --> D
C --> E
```
When two positively charged nuclei are brought close to one another, the repulsive Coulomb force creates a steep barrier that requires an initial investment of input energy to overcome. The output energy liberated from the rearrangement of nuclear binding energy must be substantial enough to justify that input and leave a net surplus to generate electricity. The second critical metric is the reaction cross section—the probability that two particles will actually undergo fusion upon collision rather than simply deflecting. Because nuclear and quantum effects dictate the shape of cross-section curves, determining them requires an entire scientific discipline of empirical nuclear measurements.
An ideal fusion fuel must satisfy three distinct criteria:
- **Low input energy:** Lowering the kinetic energy barrier makes the process technologically easier to attain and far more economical to sustain.
- **High probability of fusion:** A higher cross section maximizes the reaction rate when particles interact.
- **High output energy:** Maximizing the energy yield per reaction generates enough net power to convert efficiently into commercial electricity.
---
[11:14](https://www.youtube.com/watch?v=L0KuAx1COEk&t=674s) Narrowing Down the 3,000 Nuclides to D-T
Looking across the roughly 3,000 nuclides known in the universe, an evaluation of fusion practicality eliminates nearly every option. As soon as elements heavier than hydrogen, helium, or lithium—such as beryllium and carbon—are considered, the Coulomb barrier becomes far too high to overcome under any feasible operating conditions. Furthermore, examining the reaction cross sections reveals that the vast majority of interactions have probabilities so infinitesimally small that they are ruled out immediately. Most remaining combinations offer marginal energy outputs and low cross sections that rule them out for near-term practical use with known technology.
This aggressive winnowing process leaves essentially three viable reactions, alongside a fourth marginal fuel often pursued by alternative concepts: proton-boron 11 ($\text{p-}^{11}\text{B}$). The primary hurdle across all candidates is what Hartwig terms the "bogeyman of fusion"—Coulomb scattering. The probability that two colliding charged particles will merely deflect off each other via electromagnetic repulsion is vastly greater than the probability that they will fuse.
Zach Hartwig illustrates this dynamic with a billiard table analogy:
> “If I have no way to confine these particles, they're simply going to scatter and they're lost, right? I think of a billiard ball table with no bumpers around the edge, you know, you throw a bunch of billiard balls together, they might knock into each other, they're never going to go in the pockets cuz the probability is low, they just fall off the table.”
However, with an effective confinement scheme acting like bumpers on the table, particles can bounce off one another indefinitely until they eventually fuse and release their stored energy.
---
[14:07](https://www.youtube.com/watch?v=L0KuAx1COEk&t=847s) Plasmas, Temperatures, and Fuel Rankings
Confining particles at the kinetic energies necessary to achieve a significant fusion cross section naturally creates a plasma—an ionized gas composed of free electrons and positive ions behaving collectively as two interpenetrating fluids. In such a system, particle behavior cannot be tracked as individual trajectories; it must be treated collectively, where the average kinetic energy is measured as temperature. Moving from energy units to temperature, $15\text{ keV}$ corresponds to approximately 175 million degrees (Kelvin, Celsius, or Fahrenheit being practically interchangeable at such extreme magnitudes).
When ranked by reaction rate per unit volume per second and operational temperature, deuterium-tritium (D-T) emerges as the undisputed frontrunner:
As Hartwig emphasizes regarding the immense operational challenge of alternative fuels:
> “You know, by the time you're doing proton boron 11 fusion, you better have a plasma which is close to 2 billion degrees. This is a technologically very, very challenging.”
---
[16:01](https://www.youtube.com/watch?v=L0KuAx1COEk&t=961s) Energy Conversion Pathways and the Role of Neutrons
The choice of fuel determines the physical form in which the released binding energy emerges, dictating the engineering required to convert that energy into electricity. In a D-T reaction, the energy is partitioned into the kinetic energy of an alpha particle (a helium nucleus, $^4\text{He}$) and a high-energy neutron ($n$):
```mermaid
flowchart TD
accTitle: D-T Energy Conversion Cycles
accDescr: Flowchart illustrating the thermal and breeding cycles of neutrons and alpha particles from D-T fusion.
subgraph Products["D-T Fusion Products"]
N["Neutrons (Kinetic Energy)"]
A["Alpha Particles (Kinetic Energy)"]
end
subgraph NeutronPath["Neutron Loop"]
N --> B["Slowing Down in Blanket<br/>(Converts Kinetic Energy to Heat)"]
B --> C["Steam Generation & Turbines<br/>(Electricity)"]
N -.->|"Tritium Breeding"| D["Lithium Blanket Interaction<br/>(Generates New Tritium Fuel)"]
end
subgraph AlphaPath["Alpha Loop"]
A --> E["In-Plasma Collisions<br/>(Self-Heating / Keeps Plasma Hot)"]
E --> F["Bremsstrahlung / X-rays<br/>(Radiative Heat Loss)"]
F --> G["X-ray Heat Capture / Turbines"]
end
```
The uncharged neutron escapes magnetic confinement easily, penetrating into surrounding blanket materials where it rattles around, slows down, and converts its kinetic energy into thermal heat. This heat boils water into steam to drive a conventional turbine generator and produce electricity. Crucially, the neutron offers an elegant secondary advantage: it can be captured by lithium in the blanket to breed new tritium, enabling the reactor to sustain its own fuel cycle.
Meanwhile, the charged alpha particles remain confined by magnetic fields. Born at energies much higher than the surrounding bulk plasma, they collide with other ions and electrons, redepositing their energy to keep the reaction self-sustaining and hot. As electrons interact with ions and decelerate, they emit X-ray radiation (bremsstrahlung).
In aneutronic fuel cycles (such as $\text{p-}^{11}\text{B}$), no high-energy neutrons are available for thermal blankets. Instead, plant operators must rely on capturing intense X-ray radiation as heat, or turn to direct energy conversion—an advanced, currently speculative technology that attempts to extract electricity directly from the kinetic energy of moving charged particles without an intermediary thermal cycle.
[17:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1074s) Viable Fusion Fuels and Evaluating Reactor Claims
Reflecting on the comparative analysis of fusion fuel cycles, Zach Hartwig establishes a core rule for evaluating any fusion concept: “Rule number one is fuel choice fundamentally sets the difficulty of any approach to fusion energy.” While deuterium-tritium (DT) is the easiest reaction to ignite, deuterium-deuterium (DD) and deuterium–helium-3 ($\text{D-}^3\text{He}$) are progressively harder by orders of magnitude, and proton–boron-11 ($\text{p-}^{11}\text{B}$) may well be physically infeasible for net power production. Other theoretical fuels are simply non-starters.
Whenever one encounters sensational fusion concepts promoted in popular media outlets like *Time* magazine or general newspapers, Hartwig emphasizes that the public and scientific community must interrogate the underlying claims with specific technical questions:
- What specific fusion fuel cycle does the concept plan to use?
- Does the design team honestly acknowledge the unique physical and engineering hurdles associated with that specific fuel?
- How is the released fusion energy converted into grid electricity?
- If aggressive commercial reactor claims are being made today, has the required energy-conversion technology actually been demonstrated at scale?
---
[18:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1119s) The Lawson Criterion: Campfires and Plasma Burning Conditions
Turning to the second fundamental question—what physical conditions are required to achieve net fusion energy—Hartwig notes that although no human has ever produced a sustained self-burning plasma, everyone understands the basic physics of building a campfire. Making wood burn requires three complementary conditions: supplying enough wood without choking the flame, raising the fuel to ignition temperature (which fails if the wood is soaked, because input heat is wasted boiling water into steam), and arranging the logs so the structure confines heat to continuously drive the combustion forward.
```mermaid
flowchart LR
accTitle: The Lawson Criterion Analogy
accDescr: Comparison between the requirements for burning a wood fire and burning a fusion plasma.
subgraph WoodFire["Campfire Analogy"]
W1["Sufficient Dry Wood"]
W2["Ignition Temperature"]
W3["Log Arrangement (Traps Heat)"]
end
subgraph PlasmaBurn["Burning Plasma (Lawson)"]
P1["Plasma Density (n)"]
P2["Plasma Temperature (T)"]
P3["Energy Confinement Time (tau_E)"]
end
W1 -.-> P1
W2 -.-> P2
W3 -.-> P3
```
In plasma physics, the requirements are identical. Achieving net fusion demands an optimal particle density ($n$), a high ion temperature ($T$) to maximize fusion reactivity, and an effective confinement scheme that limits how rapidly thermal energy leaks away. This confinement quality is quantified by the energy confinement time ($\tau_E$). Derived by British engineer and physicist J. D. Lawson in 1955, the *Lawson criterion* dictates the product of these three parameters ($n$, $T$, and $\tau_E$) necessary for ignition or net power.
This condition is visualized on a two-dimensional parameter space plotting the Lawson parameter—the product of density and confinement time ($n\tau_E$)—against plasma temperature ($T$). The space spans multiple decades in both axes and is crossed by contours of the plasma energy gain factor, $Q$:
$$Q = \frac{\text{Fusion Energy Output}}{\text{Energy Input}}$$
To put net electricity onto the power grid, a reactor must operate at a high $Q$, strictly greater than 1, to generate more energy than it consumes.
---
[20:47](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1247s) Crossing the Parameter Space and the "Theme Park" Viability Metric
For the past sixty years, fusion energy research has centered on selecting a confinement concept, starting near the bottom of the parameter space, and attempting to move diagonally into the high-temperature, high-$n\tau_E$ regime in the top-right corner. Hartwig emphasizes how deceptively punishing this journey is in practice: navigating across several orders of magnitude alters the fundamental behavior of the physical system, unmasking "unknown unknowns." At each step, latent plasma instabilities can arise and degrade confinement, leaking heat and destroying a concept's viability. Some instabilities can be managed through decades of engineering and substantial capital, while others prove insurmountable and force the abandonment of entire concepts.
```mermaid
flowchart TD
accTitle: Fusion Concept Readiness Tiers
accDescr: Three color-coded readiness regions for fusion concepts based on physics proximity to burning plasma conditions.
subgraph RedZone["Red Zone: Not Tall Enough to Ride"]
R1["Fundamental physics not viable for net energy"]
R2["Q < 0.1: Early physics experiment"]
R3["Reactor claims premature and exploitative"]
end
subgraph YellowZone["Yellow Zone: Accompanied by an Adult"]
Y1["Physics approaching demonstration"]
Y2["Q ~ 0.1 to near-breakeven"]
Y3["Reactor engineering design is justified"]
end
subgraph GreenZone["Green Zone: Tall Enough to Ride"]
G1["Physics unquestionably demonstrated"]
G2["High Q achieved"]
G3["Mature reactor engineering / power to grid"]
end
RedZone --> YellowZone --> GreenZone
```
To categorize where concepts stand, Hartwig compares the parameter space to the height restrictions at an amusement park:
1. **Red Zone ("You are not tall enough to ride"):** The fundamental physics is not yet proven viable for energy production. Designing reactor components at this stage puts the cart before the horse. As Hartwig cautions, “certainly grandiose reactor claims that you're going to come up with watts in the grid in 3 years are exploitative of people who don't know enough to understand even, you know, the physics behind what you're trying to do.”
2. **Yellow Zone ("Must be accompanied by an adult"):** Concepts that approach or achieve an energy gain around $Q \sim 0.1$. The core physics is nearing quantitative demonstration, meaning the particle losses, heat flux, and operational bounds are reasonably understood. Here, engineering design studies and reactor projections become legitimate.
3. **Green Zone ("Tall enough to ride"):** Burning-plasma physics is unquestionably proven, reactor engineering is mature, and the technology is ready to generate net electric power for the grid.
Hartwig stresses that proximity to burning plasma conditions in the top-right corner is the ultimate arbiter of any fusion approach. Anyone evaluating a novel startup or machine must demand hard figures: What is the achieved plasma pressure? What is the ion temperature? What is the measured energy confinement time? What $Q$ has actually been attained on existing hardware? How many orders of magnitude in extrapolation separate present results from reactor-relevant operation? And does plasma turbulence destroy heat confinement?
---
[24:28](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1468s) Evaluating Confinement Approaches: From Taxonomy to Cold Fusion
Moving to the third guiding question—what fusion approaches exist and how to evaluate them—Hartwig references a comprehensive taxonomy hand-sketched by physicist Simon Woodruff around 2002. This landscape encompasses magnetic confinement fusion (MFE), inertial confinement fusion (IFE), open magnetic field line devices (mirrors), and alternative pinch configurations.
Before examining mainstream approaches, Hartwig addresses cold fusion, rebranded as Low Energy Nuclear Reactions (LENR):
- **Confinement Mechanism:** Completely unknown.
- **Experimental Status:** A surprising number of active fringe efforts continue worldwide.
- **Physical Validity:** Cold fusion originated with the late-1980s experiments of Martin Fleischmann and Stanley Pons. Hartwig states flatly: “it's important to point out that the proposed process cannot be rectified with any known physical model of the universe.”
- **Independent Replication:** The claims have been repeatedly debunked with zero verified replications by skeptical independent critics. Much of the early experimental debunking was carried out by researchers at the MIT Plasma Science and Fusion Center, prompting historic headlines such as the *Boston Herald* reporting an MIT "bombshell" disproving cold fusion.
- **Scientific Classification:** Cold fusion is widely analyzed within science and technology studies (STS) as a textbook case of "pathological science"—a pursuit that adherents cling to despite definitive empirical disproof.
- **Performance Rating:** Because zero cannot be plotted on a logarithmic scale, cold fusion's performance is classified strictly as "NaN" (Not a Number) on the Lawson plot.
Shifting from pseudoscience back to basic physics, Hartwig turns to nature's proven method: gravitational confinement. The universe achieves sustained fusion inside stars like the Sun, where an immense mass of plasma creates the self-gravitating pressures and core temperatures necessary to fuse hydrogen.
[27:03](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1623s) Gravitational and Explosive Fusion: The Sun and the Hydrogen Bomb
In stars, fusion progressively builds heavier elements from helium to lithium, beryllium, and beyond. This creates an ongoing contest: the burning plasma exerts thermal pressure and wants to expand outward like a balloon, while gravity pulls inward. The result is a transient equilibrium—though “transient” on a stellar scale translates to billions of years—allowing the Sun to produce colossal amounts of power. However, Zach Hartwig emphasizes that the Sun achieves this with a surprisingly tiny volumetric power density of only about a third of a watt per cubic meter, roughly comparable to an ordinary compost pile; it simply produces immense total power because its volume is vast. Replicating gravitational confinement on or near Earth is technologically unviable, even though the Sun clearly sits far up the performance curve producing net energy.
```mermaid
flowchart TD
accTitle: Competing Forces in Stellar Equilibrium
accDescr: Diagram illustrating the balance between plasma thermal pressure expanding outward and gravitational force contracting inward in stars.
subgraph Core["Stellar Fusion Core"]
direction TB
FUS["Nuclear Fusion\n(He -> Li -> Be...)"]
end
subgraph Forces["Equilibrium Forces"]
direction LR
EXP["Plasma Pressure\n(Outward Expansion)"] <---> GRAV["Gravitational Force\n(Inward Contraction)"]
end
FUS --> EXP
```
Turning to human technology, Hartwig examines the first terrestrial method that achieved net energy gain: nuclear weapons. A hydrogen bomb uses a primary fission stage (uranium or plutonium) to trigger an implosion that releases intense X-rays, which then violently compress a secondary fusion capsule to create the conditions for net energy gain. As Hartwig notes, “To date, this is the only successful net fusion energy gain on Earth.” Yet while thermonuclear explosions indisputably satisfy the physical criterion for net gain, doing so in an uncontrolled blast is impractical for commercial power production, as he quips: “There goes the power plant.”
[28:47](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1727s) Inertial Confinement Fusion: Crushing Fuel with Lasers
The concept of crushing fuel remains compelling even if bombs are unsuited for the power grid, leading to inertial confinement fusion (ICF). In this scheme, high-energy laser systems—such as those at the National Ignition Facility (NIF)—compress a tiny fuel capsule to reach fusion conditions before the fuel can blow apart. Hartwig calls NIF an “incredibly impressive experiment” that has produced outstanding physics results, estimating its scientific gain $Q$ at roughly 0.1, putting it respectably close to the break-even threshold.
```mermaid
flowchart LR
accTitle: Challenges in Laser Inertial Confinement Fusion
accDescr: Sequence showing laser drivers compressing fuel capsules and the engineering hurdles preventing continuous energy generation.
subgraph Driver["Driver Pulse"]
LAS["High-Power Lasers"] --> DRV["Target Implosion"]
end
subgraph CoreProcess["Fusion Event"]
DRV --> EXP["Capsule Disruption\n& Energy Release"]
end
subgraph EngineeringBottlenecks["Commercial Engineering Hurdles"]
DAM["Hardware Damage\nEach Shot"]
REP["Repetition Rate Deficit\n(~1/day vs. ~1/sec)"]
EFF["Very Low Driver\nWall-Plug Efficiency"]
end
EXP -.-> EngineeringBottlenecks
```
Despite these physics achievements, Hartwig points out severe engineering barriers that make near-term energy production via laser ICF difficult to envision:
- **Hardware destruction:** Significant interior components of the target chamber are destroyed during each capsule implosion.
- **Repetition rate shortfall:** Current experimental facilities fire approximately once per day, whereas an operational power plant would require roughly one shot per second—a factor of 100,000 increase in firing rate.
- **Low energy efficiency:** The overall efficiency of converting grid electricity into laser energy and coupling that power into the compressed target remains extremely low.
[30:24](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1824s) Electrostatic Approaches: Linear Accelerators and Fusors
When mechanical compression and gravity fall short, another option is the electric force. One can imagine using electric fields to accelerate deuterium and tritium ions directly into a target or into opposing beams. Hartwig recalls the persistent “bogeyman of fusion”—Coulomb scattering. Because the cross-section for electrostatic scattering is orders of magnitude larger than the fusion cross-section, accelerated particles scatter off one another and dissipate their directed energy long before sufficient fusion reactions can take place. Even so, fusion has always been “a field full of optimists,” prompting experiments like Migma IV. While these devices generated high-energy particles, their particle density and energy confinement times were so poor that they fail to even appear on standard Lawson parameter plots.
```mermaid
flowchart TD
accTitle: Loss Mechanisms in Inertial Electrostatic Confinement Fusors
accDescr: Diagram mapping the central potential well in a fusor to its various inherent energy loss pathways.
WELL["Spherical Electrostatic Well\n(Accelerates central ions)"]
WELL --> COL["Coulomb Collisions\n(Deflection before fusion)"]
WELL --> ESC["Ion Escape\n(Loss out of well)"]
WELL --> RAD["Bremsstrahlung Radiation\n(Photon emission losses)"]
WELL --> COND["Conduction Losses\n(Grid/hardware contact)"]
```
To circumvent linear beam limits, researchers constructed three-dimensional, spherical particle accelerators known as inertial electrostatic confinement (IEC) devices, or fusors. By establishing an electrostatic potential well inside a spherical chamber, ions fall toward the center, rattle back and forth, and occasionally collide. Yet fusors suffer from identical underlying physics: particles scatter out of the well, hit physical grid wires (conduction losses), and radiate their energy away via Bremsstrahlung photon emission. Hartwig emphasizes that while fusors are relatively simple and serve as fantastic educational projects—frequently featured in headlines when ambitious students produce detected fusion neutrons at home—they are fundamentally unsuitable for net energy production: “Turns out to be a poor way to make energy.”
[32:28](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1948s) Open Magnetic Confinement: Magnetic Mirrors and Squeezed Fields
With gravitational and electrostatic approaches exhausted, researchers turned to magnetic fields. Charged ions spiral around magnetic field lines, confining them in two dimensions perpendicular to the field while leaving them free to travel along the lines. To prevent particles from escaping the ends of a linear tube, early pioneers designed the “magnetic mirror,” which constricted the field at both ends with high-field magnetic coils. The theoretical concept relied on the magnetic mirror force reflecting spiraling ions back into the central chamber. In practice, however, these machines “just hemorrhaged leakage out of the ends.” Decades of engineering clever magnetic plugs revealed that closing the end losses would require a machine several kilometers long, rendering the concept economically unfeasible. Modern variants—such as the design pursued by Lockheed Martin—remain four to six orders of magnitude below the plasma performance required for energy relevance.
```mermaid
flowchart LR
accTitle: Linear Magnetic Mirror Confinement and Loss
accDescr: Conceptual diagram of a simple magnetic mirror showing particle spiraling and end loss leakage.
subgraph FieldConfig["Mirror Geometry"]
C1["High-Field Throat\n(Coil 1)"] --- MID["Lower-Field Midplane\n(Trapped spiral trajectories)"] --- C2["High-Field Throat\n(Coil 2)"]
end
MID -->|"Scattering into loss cone"| LEAK["End Hemorrhage\n(Unconfined losses)"]
```
Researchers also explored dynamic approaches that squeeze magnetic fields, such as magnetized target fusion (MTF)—a method pursued in Canada by the private firm General Fusion. While valuable for researching plasma physics and potentially serving as high-flux neutron or X-ray sources, these configurations struggle with rapid plasma cooling and severe plasma instabilities that disrupt confinement before reaching net energy conditions.
In another attempt to overcome end losses without closing the field lines smoothly, researchers attempted to link multiple mirrors together end-to-end in a closed ring. This “shotgun approach” resulted in experimental machines such as the Elmo Bumpy Torus in the 1970s and 1980s—which Hartwig celebrates as “by far, one of the best fusion names of all time.” However, breaking toroidal symmetry introduced severe drift losses that doomed the concept.
[35:51](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2151s) Self-Organized and Helical Magnetic Concepts: FRCs and Spheromaks
Reflecting on earlier open and alternative concepts such as the Elmo Bumpy Torus, Hartwig notes that while they exhibited fascinating plasma physics, they ultimately suffered from severe instabilities, restricted temperatures, and poor energy confinement. Attempting to plot them on the Lawson criterion parameter space reveals that they simply do not fit into the regime required for net power. Nonetheless, those early experiments imparted a fundamental physics insight: plasma symmetry is essential. To build on that lesson, researchers explored whether magnetic field lines could be twisted around by exploiting the self-organization of the plasma itself.
This line of inquiry led to alternative toroidal configurations like the field-reversed configuration (FRC) and the spheromak. In these concepts, internal plasma currents generate the necessary twisting magnetic field rather than relying entirely on complex external structures. A prominent modern commercial effort pursuing this route is Tri Alpha Energy, which has attracted substantial press attention and built sophisticated experimental hardware. Yet, as Hartwig points out, these self-organized concepts are plagued by inherently transient behavior and short-timescale instabilities. Despite decades of study across many laboratories, they remain many orders of magnitude below the triple product conditions necessary for net fusion energy. Nevertheless, they demonstrated a pivotal physical principle: helical confinement—twisting the magnetic field lines—effectively tames several virulent plasma instabilities.
- **Field-Reversed Configurations (FRCs) and Spheromaks:** Rely on internal plasma self-organization to twist magnetic fields toroidally, but suffer from severe short-lived transients.
- **Tri Alpha Energy:** High-profile commercial venture pursuing advanced beam-driven FRC concepts.
- **Lawson Distance:** Self-organized concepts remain many orders of magnitude below energy-relevant fusion conditions.
> “So, again, very interesting plasma physics, but if we were to go back to our space, and we look at the many types of FRCs, spheromaks, and even Tri Alpha, they are many orders of magnitude away from where they need to be.”
---
[37:24](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2244s) External Coils: The Stellarator's Resurgence
Instead of relying on the plasma to organize its own field, an alternative approach applies helical twist entirely from the outside using external electromagnetic coils: the stellarator. One of the earliest magnetic fusion concepts ever conceived, the stellarator eliminates the need to drive a large toroidal net current through the plasma, but it does so at the price of extreme geometric complexity. The magnetic coils must be sculpted into intricate, non-planar, three-dimensional shapes that twist the plasma into a ribbon-like geometry.
Hartwig highlights the Wendelstein 7-X facility in Greifswald, Germany, as one of the most exciting and ambitious fusion experiments operating today. Photographs of the device reveal human technicians dwarfed by the labyrinth of giant, twisted superconducting magnets that encircle the vacuum vessel. Unlike most alternative concepts, stellarators have successfully marched out of the bottom-left corner of the Lawson performance space, demonstrating genuine confinement improvements. While formidable engineering hurdles and high manufacturing complexities remain, their steady progress makes them a critical technology to track over the coming decade.
- **Wendelstein 7-X (Germany):** World-leading advanced stellarator utilizing highly complex, non-planar superconducting magnetic coils.
- **Performance Trajectory:** First non-tokamak magnetic configuration to make meaningful progress upward toward energy-relevant confinement regimes.
- **Engineering Trade-off:** Steady-state operational advantages balanced against extreme manufacturing tolerances and complex coil geometries.
> “And they've actually done pretty well, right? This is the first sort of configuration of devices that we've seen sort of march out of this bottom left corner and make some progress towards the right conditions.”
---
[38:17](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2297s) The Tokamak's Historic Supremacy
The final configuration examined is the tokamak, which combines external toroidal magnetic coils with an internally driven plasma current to produce the requisite helical twisting field. When Soviet physicists first operated the tokamak in the late 1960s, it immediately outperformed every competing fusion concept in existence. Western scientists traveled to the USSR to verify the results with independent laser diagnostics, and once the data was confirmed, the tokamak rapidly became the preeminent workhorse of global fusion research.
A plot of ion temperature versus the Lawson criterion reveals that tokamaks pepper the center of the parameter space, coming tantalizingly close to the break-even threshold ($Q = 1$). The fusion community has accumulated unmatched experimental and operational expertise with the architecture, having constructed roughly 170 tokamaks worldwide, including six successive machines built at MIT. It remains the singular magnetic confinement architecture that has reached the doorstep of net fusion energy.
```mermaid
flowchart TD
accTitle: Evolution of Magnetic Confinement Concepts
accDescr: Diagram illustrating the progression from open and self-organized magnetic configurations to stellarators and tokamaks.
subgraph Advanced["High Confinement Helical Fields"]
C["Stellarators (W7-X)<br/>(External 3D Coils)"]
D["Tokamaks (170 Built Globally)<br/>(External Coils + Plasma Current)"]
end
A -->|"Need symmetry & helical twist"| B
B -->|"Replace self-organization with external coils"| C
B -->|"Combine external field with induced current"| D
D -->|"Approaching Q >= 1"| E["Net Energy / Burning Plasma"]
```
- **Global Tokamak Fleet:** Over 170 tokamaks constructed worldwide across six decades of research.
- **MIT Experience:** Six distinct tokamaks designed, built, and operated on the MIT campus.
- **Performance Benchmark:** Tokamaks achieved an energy gain factor reaching $Q \approx 0.65$, far exceeding all competing designs.
> “And when the first tokamak was basically sort of created, it immediately outperformed almost any fusion concept to date.”
---
[39:32](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2372s) Evaluating Fusion Concepts: ITER and the Non-Tokamak Gap
Summarizing the overall landscape, Hartwig points out that every non-tokamak approach (with the partial exception of stellarators) remains clustered in the lower-left corner of the Lawson triple product diagram—in his words, “not tall enough to ride.” Startups or laboratory programs claiming they will deliver a commercial power plant within two years are fundamentally detached from the physical reality of their confinement parameters. Decades of overcoming instabilities in tokamaks cost billions of dollars, but those hard-won physics solutions are entirely unique to the tokamak; they do not rescue or validate alternative concepts that still must solve their own stability issues from scratch.
By contrast, the international community has placed its collective bet on the tokamak through ITER, the colossal international experiment under construction in Cadarache, France. As the largest science experiment on Earth, ITER involves the collaboration of the world's leading industrial nations. A human scale figure inside the cross-section is reduced to a tiny smudge of pixels against the multi-story vacuum vessel and giant magnet coils. ITER represents an undeniable statement of scientific consensus: when nations must commit real capital to net electricity production from fusion, the tokamak is the only concept deemed ready.
- **Non-Tokamak Deficit:** Non-tokamak devices trail tokamak plasma performance parameters by factors of $10^4$ to $10^6$.
- **ITER Project:** Massive multi-national experiment under construction to demonstrate $Q \ge 10$ ($500\text{ MW}$ fusion power from $50\text{ MW}$ input).
- **Physics Non-Transferability:** Solved tokamak instabilities do not resolve the unaddressed plasma physics problems of alternative configurations.
> “One way to look at ITER is to say that ITER is a $50 billion statement of conviction about the tokamak by the world's leading scientific nations and institutions that the tokamak and the tokamak alone has achieved sufficient performance and it's ready to tackle producing electricity from net fusion energy.”
> “It's an important point, however, to say that the tokamak's physics achievements do not validate, nor do they elevate non-tokamak claims to be nearing the production of fusion energy.”
---
[42:03](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2523s) The Three Rules of Fusion and Critical Questions
Hartwig returns to the aeronautics analogy introduced earlier in the lecture: just as any viable airplane concept must account for its wings, any fusion concept must be interrogated on its Lawson triple product: temperature ($T$), density ($n$), and energy confinement time ($\tau_E$). When assessing ambitious commercial claims or novel reactor pitches—including fanciful cultural touchstones like the DeLorean’s “Mr. Fusion” from *Back to the Future*—evaluators must demand the hard parameters rather than rhetoric. Non-tokamak alternatives trail the tokamak's record performance ($Q \approx 0.65$) by factors between 10,000 and 1,000,000 in this key metric.
To rigorously scrutinize any proposed fusion concept, Hartwig outlines essential questions: Has this approach been tried and abandoned before? If so, why was it dropped, and what has changed to justify reviving it? How far is the design extrapolating in density, temperature, and confinement time from existing experimental data? And by what credible mechanism, over what timeframe, and at what financial cost does the team expect to bridge a multi-order-of-magnitude performance gap?
He recaps the three foundational rules established throughout the lecture:
1. **Fuel choice fundamentally sets the difficulty:** Deuterium-tritium is the only viable near-term fuel.
2. **Proximity to a burning plasma is the ultimate arbiter:** Achieving self-sustaining alpha heating governs practical viability.
3. **Many approaches have been tried, but only the tokamak has delivered:** Tokamaks stand alone in reaching energy-relevant parameters.
> “Only the tokamak has demonstrated the right level of performance. So, the tokamak, you know, with a maximum performance of about 0.65, you know, leads by a lot.”
> “How do you propose achieving, you know, this factor of 10,000 to a million times performance increase that you need to be energy relevant? How long is that going to take? At what cost?”
---
[43:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2619s) The Tokamak's Mid-1990s Stagnation: Physics vs. Size
Despite the tokamak's decisive leadership, Hartwig cautions that all is not well within mainstream tokamak research. Plotting the fusion triple product ($n \cdot T \cdot \tau_E$) over historical time reveals that tokamaks achieved exponential, Moore's Law-like progress through the 1970s and 1980s, climbing rapidly toward reactor requirements. However, around the mid-1990s, progress hit an abrupt plateau and ceased entirely.
Hartwig asks the pivotal question: Did tokamak progress halt because the concept hit a hard physical performance ceiling, an insurmountable plasma instability, or some unforeseen physical showstopper? The answer is an emphatic no. Tokamak physics did not saturate because of physics limits; it stalled purely because of physical size. To increase confinement further under the conventional magnet technology of the era, machines had to become enormously large, culminating in JET (the Joint European Torus) reaching peak performance before the global effort was forced to build the gargantuan, slow-moving ITER.
- **Historical Progression:** Rapid exponential growth in fusion triple product from the late 1960s through the mid-1990s.
- **Mid-1990s Plateau:** Progress in the Lawson triple product parameter halted after peak performance runs on machines like the Joint European Torus (JET).
- **Root Cause of Stagnation:** Physical dimensions and capital scale required by conventional copper and low-temperature superconducting magnets, rather than fundamental plasma physics limits.
> “Physics performance ceased in the tokamak due to size. It did not saturate due to any physics problem.”
[44:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2694s) The ITER Pathway and the Standard Tokamak Critique
The historical trajectory of tokamak development reached a landmark with JET in 1997, but the subsequent step on the global roadmap is ITER. Shown to scale, a human figure shrinks to just a few pixels against ITER's colossal cross-section. ITER's peak performance is not projected until roughly 2035 to 2040, creating a sharp dogleg on performance progress charts driven almost entirely by sheer physical size. Under this traditional pathway, today's operating facilities lead to ITER operation in the 2030s, followed by intermediate steps such as a Fusion Nuclear Science Facility, and finally a European demonstration power plant (DEMO) sometime after 2050.
```mermaid
flowchart LR
accTitle: Traditional ITER Fusion Energy Roadmap
accDescr: Linear progression from current tokamaks to ITER, testing facilities, and DEMO across several decades.
subgraph Present["Present Day"]
A["Operating Tokamaks\n(JET, DIII-D, etc.)"]
end
subgraph LongTerm["Post-2050"]
D["DEMO Reactor\n(Commercial Prototype)"]
end
A --> B
B --> C
C --> D
```
This extended timeline fuels the widespread criticism that tokamaks are fundamentally unsuitable for addressing near-term energy and climate challenges. Zach Hartwig directly validates the frustration behind this critique while cautioning against throwing away the underlying physics: “Tokamaks are too big, they are too complex, and they are too slow to be relevant to any issue in the near term on climate change and energy production issues.” He emphasizes that this protracted development was not an inherent flaw of plasma physics, but rather the result of past technological compromises and the massive geopolitical bureaucracy required to coordinate an experiment on a global scale. The immense size and cost of ITER do not mean the tokamak will fail, nor do they justify abandoning the concept altogether.
- JET demonstrated peak performance in 1997; ITER peak performance is slated for roughly 2035–2040.
- European DEMO reactor deployment is projected only after 2050 on the traditional international track.
- Organizational complexity and geopolitical overhead are direct consequences of building at the multi-national gigawatt scale.
---
[46:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2799s) The Leverage of Magnetic Field: $B^2$, $B^3$, and $B^4$ Scaling
The physical dimensions of a tokamak are fundamentally dictated by the strength of its magnetic field ($B$). Magnetic fields provide confinement by forcing charged plasma particles into tight helical orbits around magnetic field lines; higher magnetic field strength shrinks the gyroradius, leading to vastly superior insulation and better energy confinement. Stronger fields also supply stabilizing magnetic pressure to resist the thermal expansion of the hot plasma, scaling directly with the square of the field ($B^2$).
```mermaid
flowchart TD
accTitle: Magnetic Field Scaling Laws in Tokamaks
accDescr: How increasing magnetic field strength scales confinement, magnetic pressure, fusion gain, and power density.
B["Magnetic Field (B)"]
B -->|"Linear to Tight Gyroradius"| Conf["Confinement & Insulation"]
B -->|"B² Scaling"| Press["Magnetic Pressure\n(Plasma Stability)"]
B -->|"B³ Scaling"| Q["Fusion Energy Gain (Q)"]
B -->|"B⁴ Scaling"| Power["Fusion Power Density\n(Economic Viability)"]
```
Crucially, the fusion reactivity per unit volume depends exceptionally strongly on the magnetic field. Fusion energy gain ($Q$) scales approximately as $B^3$, while the volumetric fusion power density—how much power is generated per cubic meter of core plasma, which governs economic viability—scales as the fourth power of the magnetic field ($B^4$). When plotting device size versus magnetic field strength along contours of constant fusion gain ($Q$), one sees that achieving net energy with weak magnetic fields requires global-scale, multi-nation consortia like ITER. By contrast, moving toward higher magnetic fields allows the exact same net-energy contours to be reached at the national laboratory, university, or private company scale.
- Confinement and insulation improve as particles gyrate more tightly around higher-$B$ field lines.
- Magnetic pressure scales as $B^2$, balancing plasma density and temperature.
- Fusion energy gain ($Q$) scales as $B^3$.
- Fusion volumetric power density scales as $B^4$, directly impacting plant capital cost and size.
---
[48:49](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2929s) Technological Locks and the High-Temperature Superconductor Revolution
ITER's enormous footprint was not chosen arbitrarily; it was locked in by the state of magnet technology when the machine's baseline was frozen. Superconducting magnets are essential for steady-state fusion because they conduct electricity without resistive losses, requiring power only for cryogenic cooling. However, the conventional low-temperature superconductors (LTS) available at the time—such as niobium-tin ($\text{Nb}_3\text{Sn}$)—could sustain peak magnetic fields on axis of only around 5.5 Tesla before losing their superconducting state and quenching. Hartwig paraphrases the ITER engineering design philosophy: “We understand the physics of tokamaks. We know the importance of magnetic field. We will go to the highest magnetic field the technology lets us go to, and then we will simply have to fix our size to get the performance up.” Bound to roughly 5.5 Tesla, the designers had no choice but to scale the machine's physical size up to a massive 840 cubic meters of plasma to achieve a target gain of $Q = 10$.
| Parameter | Low-Temperature Superconductors (LTS / ITER era) | High-Temperature Superconductors (HTS / REBCO) |
|:--- |:--- |:--- |
| **Operating Regimes** | Extremely narrow critical temperature, field, and current space | Radially expanded critical surface across temperature, field, and current |
| **Max On-Axis Tokamak Field** | $\sim 5.5\text{ T}$ | Potential for $9\text{ T}$ to $>12\text{ T}$ |
| **Cooling Requirements** | Liquid Helium ($\sim 4.2\text{ K}$) | Flexible cryogenic options; higher operating temperatures possible |
| **Current Density & Strength** | Lower critical current density; bulky structural footprint | Extremely high current density in thin tapes; allows compact, robust structural support |
| **Supply Chain Status** | Legacy industrial base | Transitioned from laboratory curiosity to commercially manufactured tape |
In recent years, the commercial emergence of rare-earth barium copper oxide (REBCO) high-temperature superconductors (HTS) shattered this historical magnetic field ceiling. Unlike LTS, HTS exhibits a vastly expanded operating space across critical temperature, magnetic field, and critical current density. Rather than remaining an exotic academic curiosity, HTS is now manufactured industrially in high-performance coated conductor tapes by multiple commercial suppliers. This material leap allows magnet engineers to either drastically shrink the volume of a reactor while preserving performance, or vastly increase the power output of a given envelope, all while utilizing higher-temperature cryogenics and thinner magnet coils with superior structural support.
---
[51:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3114s) Redesigning Fusion: ARC and the SPARC Initiative
Breaking the magnetic field constraint enabled researchers at MIT to fundamentally rethink tokamak architecture. In a conceptual design study led by Professor Dennis Whyte with MIT graduate students, the team conceived ARC (Affordable, Robust, Compact). ARC applied industrial HTS tapes to operate at an on-axis magnetic field of approximately 9.2 Tesla with a major radius of about 3.3 meters. By leveraging $B^4$ power scaling, ARC is designed to produce 500 megawatts of fusion power—matching ITER's output—using the exact same well-verified tokamak physics rules, but inside a vacuum vessel roughly ten times smaller by volume. This order-of-magnitude reduction slashes capital expenditure, shortens development schedules, and reduces institutional overhead.
```mermaid
flowchart TD
accTitle: High-Field Tokamak Development Sequence
accDescr: Comparison of ARC and SPARC downscaling paths driven by high-temperature superconductors.
subgraph SPARC_Design["SPARC Prototype"]
S1["Smallest Possible ARC\n(University / Startup Scale)"]
S2["Demonstrate Net Energy (Q > 1)\nWithout $50B Mega-Consortia"]
S1 --> S2
end
HTS --> ARC_Design
HTS --> SPARC_Design
```
MIT’s Plasma Science and Fusion Center took this logic one step further: if higher field enables a compact power plant, pushing magnetic field to its technological limits enables an even smaller intermediate device. This motivated the design of SPARC ("Smallest Possible ARC"). SPARC is conceived as a compact, university- or startup-scale tokamak engineered to cross the break-even threshold and produce net fusion energy ($Q > 1$) for the first time in human history. Hartwig underscores the pragmatic power of this strategy: “You don't have to organize the world's most advanced nations to build a $50 billion experiment. That's a really compelling argument for why you build high-magnetic field tokamaks.” Rather than gambling on untested plasma physics, high-field tokamaks rely on decades of established toroidal confinement data, using HTS magnet innovation to make fusion energy agile, affordable, and timely.
- ARC achieves 500 MW fusion power at a 3.3-meter major radius and 9.2 Tesla field, approximately one-tenth the volume of ITER.
- SPARC scales the ARC concept down to the smallest possible size capable of demonstrating net energy gain ($Q > 1$).
- High-field magnets shift fusion research out of $50 billion multi-governmental deadlocks and place it into academic and private innovation cycles.
[54:07](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3247s) MIT's Vision: From Alcator C-Mod to SPARC and Power on the Grid
Zach Hartwig concludes his formal presentation by summarizing the ultimate motivation behind fusion energy research: climate change, the sheer scale of global energy production needs, and the geopolitical and environmental issues surrounding where humanity gets its energy. If society requires near-term solutions, tokamaks represent the most viable foundation. However, by applying a decisive technical innovation—high-temperature superconductors—an even more compelling route opens up. Hartwig traces this direct evolution, pointing to the research tokamak located right on MIT's campus (Alcator C-Mod, which visitors can tour at 4:00 PM), scaling through a proposed compact test device called SPARC, and leading toward practical power plants.
```mermaid
flowchart LR
accTitle: MIT Tokamak Pathway to Commercial Fusion
accDescr: Three-stage progression from Alcator C-Mod to SPARC and commercial power generation on the grid.
subgraph Progression ["MIT Fusion Pathway"]
A["Alcator C-Mod<br/>(Campus scale)"] --> B["SPARC<br/>(Net-energy demonstration)"]
B --> C["Commercial Tokamak<br/>(Watts on the grid in 10-15 years)"]
end
```
Hartwig leaves his summary slide up and reminds the audience how to evaluate the confusing “panoply of fusion concepts” pitched across the industry. Just as the critical metric for an airplane is “what does your wing look like?”, in fusion the essential test is where a concept lands on a plot of the Lawson parameter versus temperature. Tokamaks are clearly orders of magnitude ahead of alternative schemes. Emphasizing the timeline and institutional focus, Hartwig states: “And if doing things on a short time scale with high confidence is what we need to do, tokamaks look pretty good. And the pathway that's being envisioned at the PSFC takes us from where we are today and goes on a short time scale to something that puts watts on the grid.”
- Time horizon: Feasibly putting net electricity on the grid within 10 to 15 years.
- Core evaluation metrics: Lawson parameter ($n\tau_E$) and plasma temperature ($T$).
- Advantage: High confidence combined with orders-of-magnitude lead over competing confinement ideas.
---
[55:50](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3350s) Q&A: Boron Neutron Capture and the Problem of Recirculating Power
An audience member opens the question-and-answer period by asking about the fourth entry on Hartwig’s earlier reaction table concerning proton capture, noting that MIT previously ran a project using boron neutron capture therapy (BNCT) to treat brain tumors. The questioner remarks that researchers solved the challenge of delivering boron into tumors and “somehow they managed to convince patients to slide under the nuclear reactor,” but wonders whether that therapy was a fusion or a fission process, and how it relates to fusion energy.
Hartwig clarifies that there is fundamentally nothing unique about a “fusion” reaction compared to other nuclear rearrangements. Whenever nuclei restructure—whether two deuterons fuse or a neutron is absorbed by a nucleus—it is broadly classified as a nuclear reaction. Neutron absorption reactions typically offer exceptionally high energy gain and have the largest interaction cross-sections known in physics. Hartwig notes that Todd Rider’s original analysis even included a separate column for neutron-driven reactions, which initially appear very attractive on paper due to their enormous cross-sections and favorable energetics.
However, the fundamental barrier to using neutron-driven cycles for net power is the extreme energetic cost of producing free neutrons in the first place. Hartwig explains: “And so, if your idea is to make neutrons, then use reactions to create electricity, you still have to feed that electricity back in to make neutrons.” This creates an insurmountable penalty in recirculating power—the fraction of generated power required just to keep the reactor operating. Because generating free neutrons requires so much energy input, any scheme relying on artificial neutron production fails to achieve net power on the electrical grid.
---
[58:28](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3508s) Q&A: The Resurrection of High-Temperature Superconductors (REBCO)
A second audience question addresses high-temperature superconductors (HTS). The attendee observes that HTS was discovered in the 1980s, followed by a major loss of excitement in the early 1990s when researchers largely abandoned them in favor of traditional low-temperature niobium superconductors (such as niobium-tin and niobium-titanium), and asks why HTS was sidelined for so long.
Hartwig explains that when rare-earth barium copper oxide (REBCO) materials were initially discovered, they were merely fragile, microscopic crystalline grains sitting on lab benches. While their ability to superconduct at elevated temperatures electrified the scientific community—sparking a landmark event dubbed the “Woodstock of physics” (which Hartwig quips was “not nearly as fun and debaucherous as the actual Woodstock”) where physicists crowded onto tables to see the demonstrations—the material was completely unsuitable for practical magnet engineering. Turning brittle, microscopic grains into continuous, flexible industrial conductors took over two decades of materials science development.
Only within the preceding 5 to 10 years did thin-film deposition techniques, such as Ion Beam Assisted Deposition (IBAD) and Rolling Assisted Biaxially Textured Substrates (RABiTS), mature enough to control crystalline grain alignment over long distances. Hartwig highlights that this manufacturing breakthrough transformed the field: “So, it's really a manufacturing innovation that took it from a benchtop curiosity to something that's an industrial-scale product.” Today, REBCO can be produced in kilometer-length ribbons resembling stainless steel tape, with steadily improving current-carrying performance.
- Material: Rare-earth barium copper oxide (REBCO).
- Historical event: The March 1987 APS meeting, known as the “Woodstock of physics.”
- Manufacturing processes: IBAD (Ion Beam Assisted Deposition) and RABiTS (Rolling Assisted Biaxially Textured Substrates).
- Modern form factor: Steel-like flexible tapes manufactured at kilometer scales.
---
[01:00:51](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3651s) Q&A: Grid Integration, Tritium Breeding, and Steady-State Confinement
An audience member asks whether major engineering bottlenecks remain between achieving high fusion gain ($Q$) inside a tokamak and actually exporting electricity to the power grid. Hartwig points out that converting neutron kinetic energy into electricity has been understood since the 1940s and 1950s. He highlights Experimental Breeder Reactor I (EBR-I) in Idaho, which famously lit four light bulbs in December 1951—the first time humanity generated electricity from nuclear fission. The basic thermal cycle of using neutrons to produce heat, boil water, drive steam, and spin a turbine is established engineering.
Nonetheless, Hartwig outlines two severe engineering hurdles unique to deuterium-tritium (D-T) fusion:
1. **Materials Radiation Damage**: The intense flux and high energy of fusion neutrons (14.1 MeV) displace atoms within structural walls and degrade magnet components over time.
2. **Net Tritium Breeding**: Tritium does not exist in meaningful quantities in nature and has a short half-life (~12.3 years). A commercial D-T reactor must incorporate lithium blankets to capture outgoing neutrons and produce at least one replacement triton per fusion event (a tritium breeding ratio greater than 1).
A follow-up question pushes Hartwig on another fundamental physics debate: steady-state operation versus pulsed operation, particularly when contrasting tokamaks with stellarators. The questioner notes that tokamaks inherently rely on transformer action (which is naturally pulsed unless complex non-inductive current drive is used), whereas stellarators are inherently steady-state devices. Hartwig acknowledges the trade-off and prepares to address the balance between engineering complexity, pulse length, and how close each concept is to commercial reality.
[01:03:21](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3801s) Tokamaks, Stellarators, and the Steady-State Frontier
Comparing magnetic confinement concepts, both tokamaks and stellarators present distinct tradeoffs, yet the primary metric remains how swiftly a device can produce net energy. Overcoming the pulsed nature of standard tokamaks transforms a perceived weakness into an engineering design opportunity. Achieving steady-state operation is a core objective for the research pathway pursued at the MIT Plasma Science and Fusion Center (PSFC). Converting existing facilities like Alcator C-Mod into compact, rapid, and cost-effective testbeds allows researchers to resolve the plasma physics of continuous current drive. At Alcator C-Mod, experiments have proven that radiofrequency (RF) waves can non-inductively drive plasma current indefinitely in steady state. While scaling these physics principles up to a power-producing regime remains to be proven, initial results indicate that steady-state tokamak operation is physically viable.
> “Because I think the instant that you have a device which makes net fusion energy, everything changes.” — Zach Hartwig
```mermaid
flowchart TD
accTitle: Pathway to Steady-State Net Fusion
accDescr: Flowchart illustrating how HTS magnets and RF current drive lead to net energy and commercial fusion.
subgraph Physics ["Core Physics Validation"]
RF["RF Wave Current Drive<br/>(Alcator C-Mod)"]
SS["Steady-State Plasma Physics"]
RF --> SS
end
subgraph MagnetTech ["Magnet Innovation"]
HTS["REBCO HTS Conductors"]
HighB["High Magnetic Fields (~20 T)"]
HTS --> HighB
end
subgraph DemonstrationPath ["Reactor Roadmap"]
SPARC["SPARC<br/>(Net Fusion Gain: Q > 1)"]
ARC["ARC Conceptual Reactor<br/>(Grid Power Plant)"]
SS --> SPARC
HighB --> SPARC
SPARC --> ARC
end
```
[01:05:11](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3911s) The ARC Conceptual Design and the SPARC Roadmap
When asked whether a formal construction budget exists for the ARC reactor, Hartwig clarifies that ARC represents an academic design exercise rather than an imminent construction project. ARC originated as an MIT nuclear science and engineering design class where students explored the implications of applying high-temperature superconductors (HTS) to tokamak architecture without being bound by conventional assumptions. The resulting concept introduced multiple solutions to longstanding nuclear engineering hurdles, demonstrating the profound scaling benefits of high magnetic fields. However, the operational roadmap requires building SPARC first to demonstrate net fusion energy gain ($Q > 1$) before finalizing the engineering configuration of a commercial power plant like ARC.
> “So, that's actually one of the interesting outcomes of ARC was that students sort of took this project and made it an interesting design study.” — Zach Hartwig
- ARC serves as a design exploration rather than a shovel-ready plant seeking capital investment.
- SPARC remains the immediate hardware demonstration targeted at achieving net fusion energy gain.
- The design features incorporated into ARC demonstrate how high-field HTS magnets radically transform reactor scale compared to large conventional designs such as ITER.
[01:07:08](https://www.youtube.com/watch?v=L0KuAx1COEk&t=4028s) Engineering Limits to Increasing Magnetic Fields
Addressing whether continuing advancements in HTS tapes could allow engineers to double the magnetic field again and shrink tokamaks even further, Hartwig outlines the hard limits imposed by nuclear physics and structural mechanics. While plasma confinement physics steadily improves as the magnetic field increases, shrinking the device eventually conflicts with the nuclear blanket requirements. Deuterium-tritium (DT) fusion releases high-energy 14.1 MeV neutrons that require a finite physical thickness of surrounding shielding and breeding material to moderate, capture, and convert their energy into usable heat. Attempting to reduce the reactor's minor and major radius down to half-meter scales removes the space necessary to capture these neutrons effectively.
Even more acute is the structural integrity of the magnet coils under immense Lorentz forces ($J \times B$). The magnetic stresses scale with the square of the magnetic field ($B^2$). Injecting higher current densities into magnet coils placed in close proximity generates immense mechanical stresses that can exceed the structural yield strength of the materials, causing the coils to deform or fracture. The mechanical limits of structural materials—such as high-strength steel jackets and structural supports—establish the true practical ceiling for magnetic field strength in a fusion power system.
> “And so, one of the key limits in a fusion device for how big you can set your magnetic field is, can you simply engineer magnets that can tolerate the structural loads of going to higher field?” — Zach Hartwig
[01:09:23](https://www.youtube.com/watch?v=L0KuAx1COEk&t=4163s) Fusion's Greatest Hurdle: Perception and Net Energy Gain
An audience member asked Hartwig to pinpoint the single most decisive problem facing fusion energy—whether scientific, engineering, or material. Hartwig emphasizes that the foremost barrier is neither an unsolved equation in plasma physics nor an unengineered component. Instead, it is the pervasive public and institutional cynicism that fusion is forever thirty years away. The pivotal milestone required to overcome this skepticism is simply placing a confirmed data point in the net energy gain regime of the plasma performance chart. Demonstrating verifiable net energy gain immediately silences the perception that fusion is an impractical dream, unlocking the capital and multi-disciplinary engineering talent necessary to deploy clean fusion power onto the commercial grid.
> “The biggest problem in fusion is perception. It's the perception that fusion's a joke, that it's not serious, that it's 30 years away and always will be, that this is not a real contributor to energy in the near term.” — Zach Hartwig
- The central mission of the MIT PSFC pathway is to place an experimental data point beyond breakeven ($Q > 1$).
- Achieving net energy changes the development trajectory from speculative research to an urgent, solvable engineering deployment effort.
- The session concludes with audience applause for the presentation.
## Notable quotes
> [01:09:23](https://www.youtube.com/watch?v=L0KuAx1COEk&t=4163s) The biggest problem in fusion is perception. It's the perception that fusion's a joke, that it's not serious, that it's 30 years away and always will be, that this is not a real contributor to energy in the near term. — Zach Hartwig
> [51:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=3114s) You don't have to organize the world's most advanced nations to build a $50 billion experiment. That's a really compelling argument for why you build high-magnetic field tokamaks. — Zach Hartwig
> [17:54](https://www.youtube.com/watch?v=L0KuAx1COEk&t=1074s) Rule number one is fuel choice fundamentally sets the difficulty of any approach to fusion energy. — Zach Hartwig
> [43:39](https://www.youtube.com/watch?v=L0KuAx1COEk&t=2619s) Physics performance ceased in the tokamak due to size. It did not saturate due to any physics problem. — Zach Hartwig
## Mentioned
MIT Plasma Science and Fusion Center (PSFC), MIT Department of Nuclear Science and Engineering (NSE), Zach Hartwig, Dan Brunner, Bob Mumgaard, Brandon Sorbom, Todd Rider, Dennis Whyte, Nuno Loureiro, David Keith, J. D. Lawson, Simon Woodruff, Martin Fleischmann, Stanley Pons, National Ignition Facility (NIF), Lawrence Livermore National Laboratory, MIT Sloan School of Management, Alcator C-Mod, SPARC, ARC, ITER, JET (Joint European Torus), Wendelstein 7-X, Tri Alpha Energy, General Fusion, Lockheed Martin, Migma IV, Elmo Bumpy Torus, Experimental Breeder Reactor I (EBR-I), REBCO, Ion Beam Assisted Deposition (IBAD), Rolling Assisted Biaxially Textured Substrates (RABiTS)
## Video description
Fusion energy and MIT's pathway for accelerated demonstration with high-magnetic field tokamaks
An introduction to the key concepts of producing clean, safe, and carbon-free electricity from magnetic fusion energy. This talk reviews the present state of fusion energy research and then introduce MIT's proposed pathway to use high-field superconducting magnets to achieve fusion energy at smaller unit size, at lower cost, and on a timescale relevant to climate change.
MIT Assistant Professor Zach Hartwig argues that the tokamak remains the only fusion architecture with demonstrated proximity to net energy, and that adopting high-temperature superconductors provides a credible pathway to bypass the multi-decade timelines of giant facilities like ITER by shrinking reactor size.
Fuel Energetics and Confinement Screening (11:14)
Hartwig frames fuel selection as the primary determinant of reactor difficulty. Among roughly 3,000 known nuclides, nuclear binding energy and Coulomb repulsion eliminate almost all options due to prohibitive ignition temperatures and low cross sections. While aneutronic alternatives such as proton-boron 11 (p-11B) eliminate neutron radiation damage, they require plasma temperatures approaching 2 billion degrees Celsius. Deuterium-tritium (D-T) remains the only practical fuel for near-term net power, igniting at roughly 15 keV (around 175 million degrees Celsius) and delivering the highest volumetric reaction rate, despite producing 14.1 MeV neutrons that necessitate blanket shielding and tritium breeding.
The Lawson Criterion and Non-Tokamak Deficits (18:39)
Net energy production requires satisfying the Lawson criterion—the triple product of plasma density (n), ion temperature (T), and energy confinement time (tau_E). Using an aviation analogy where a concept must possess functional wings rather than mere cabin amenities, Hartwig argues that many alternative concepts fail to demonstrate foundational confinement physics. Electrostatic fusors and linear accelerators suffer from dominant Coulomb scattering losses, while magnetic mirrors historically leaked plasma out of their open ends. Modern alternatives such as field-reversed configurations and spheromaks remain transient and trail tokamak performance by factors of 10,000 to 1,000,000 in triple product space. Although stellarators like Wendelstein 7-X demonstrate genuine progress with external 3D magnetic coils, roughly 170 tokamaks worldwide have historically reached closest to breakeven (Q ~ 0.65).
Magnetic Field Scaling and the HTS Revolution (46:39)
Tokamak progress plateaued in the mid-1990s following machines like JET, which Hartwig attributes to the dimensional limits of conventional magnet technology rather than underlying plasma physics ceilings. Because volumetric fusion power density scales with the fourth power of the magnetic field (B^4) and fusion gain scales as B^3, increasing magnetic field strength enables drastic reductions in machine volume. ITER's massive 840-cubic-meter core was dictated by low-temperature superconductors like Nb3Sn, which restricted on-axis fields to roughly 5.5 Tesla.
The Compact Pathway: ARC and SPARC (51:54)
Commercializing Rare-Earth Barium Copper Oxide (REBCO) high-temperature superconductor (HTS) tapes enabled MIT to design ARC, a conceptual reactor operating at approximately 9.2 Tesla. ARC produces 500 megawatts of fusion power—matching ITER's output—inside a vessel roughly ten times smaller by volume. To validate net energy without multi-billion-dollar international consortia, MIT's roadmap prioritizes SPARC, a compact intermediate prototype designed to achieve Q > 1. Hartwig notes that practical magnetic field increases are ultimately capped not by plasma physics, but by mechanical limits: magnet structures must endure Lorentz forces that scale as B^2, and D-T reactors require sufficient blanket thickness to moderate 14.1 MeV neutrons and breed replacement tritium.
Hartwig concludes that the foremost barrier to commercial fusion is not an unsolved plasma equation, but widespread institutional cynicism that net energy is perpetually decades away. Placing an empirical data point beyond breakeven (Q > 1) via a high-field device like SPARC is intended to shift fusion from speculative physics research into an urgent, deployable engineering discipline.
Source quotes
[69:23] Zach Hartwig: The biggest problem in fusion is perception. It's the perception that fusion's a joke, that it's not serious, that it's 30 years away and always will be, that this is not a real contributor to energy in the near term.
[51:54] Zach Hartwig: You don't have to organize the world's most advanced nations to build a $50 billion experiment. That's a really compelling argument for why you build high-magnetic field tokamaks.
[17:54] Zach Hartwig: Rule number one is fuel choice fundamentally sets the difficulty of any approach to fusion energy.
[43:39] Zach Hartwig: Physics performance ceased in the tokamak due to size. It did not saturate due to any physics problem.
Original transcript
[0:00] All right, well,
[0:02] I just like to welcome all of you to
[0:05] This is the first in this series of
[0:08] IAP talks in the Plasma Science and
[0:10] Fusion Center for 2017.
[0:14] For those of you interested, a little
[0:16] bit after this talk, there'll be a tour
[0:18] of our facility. That's at 175 Albany
[0:22] Street, and
[0:24] people should convene in the conference
[0:26] room in NW17 building.
[0:28] So, anybody interested in learning a
[0:30] little more about
[0:32] the laboratory.
[0:34] I also want to advertise
[0:37] a few other talks. On Friday, we're
[0:40] having two talks and a tour on the high
[0:43] energy density plasma physics program at
[0:47] MIT and at and at Lawrence Livermore,
[0:49] the National Ignition
[0:51] Facility, which is out in
[0:52] at Livermore.
[0:54] Um
[0:57] Then a couple weeks later, there's a
[0:59] talk in in sort of a something a bit
[1:02] different for us, talking about
[1:04] alternate ways of funding fusion
[1:06] research
[1:07] from a professor in the Sloan School.
[1:11] Um and then a talk on
[1:14] plasma physics and self-magnetization,
[1:16] which is very common phenomenon in the
[1:19] universe, and that'll be given by Nuno
[1:21] Loureiro.
[1:22] And then finally,
[1:24] this talk from David Keith from Harvard
[1:26] on climate control. It reads a little
[1:29] bit like something from a
[1:31] James Bond villain, I guess,
[1:34] but
[1:35] he tells me that it's it's all benign.
[1:38] Um
[1:40] Okay, let me just quickly introduce to
[1:41] our speaker today, that's Zach Hartwig.
[1:43] He was a a graduate student in nuclear
[1:46] science and engineering, graduated in
[1:48] 2014,
[1:50] I think.
[1:51] Uh was a postdoc after this and recently
[1:53] accepted an appointment as a assistant
[1:56] professor in the Department of Nuclear
[1:57] Science and Engineering. And he's going
[1:59] to give us an introduction to fusion
[2:01] energy and talk a little bit about some
[2:04] of our plans at the at the Fusion
[2:06] Center.
[2:07] Thank you.
[2:08] All right. Thanks, Martin, for the
[2:11] Mic on?
[2:14] Are we good going here?
[2:16] Is that on? Everybody That sounds on?
[2:18] Okay, good.
[2:19] Uh all right. So, I want to open by by
[2:22] thanking a few individuals. So, Dan
[2:23] Brunner, Bob Mumgaard, and Brandon
[2:25] Sorbom, uh who are colleagues at the
[2:26] PSFC, were instrumental in putting the
[2:29] talk together. Uh and then Todd Todd
[2:31] Rider, who's an alum of the NSE, he's
[2:33] actually thought very hard about many,
[2:34] many different ways to approach fusion.
[2:36] Um and some of his material he agreed to
[2:38] let me use uh for this talk. So, uh
[2:40] thanks to him.
[2:42] Okay, so
[2:43] fusion energy is a bit awkward. I'm
[2:45] actually going to stand on here and try
[2:46] not to fall off.
[2:48] Um so, this is a talk about fusion
[2:50] energy, so let's start by talking about
[2:51] energy. So, I think most of the people
[2:53] in this room are familiar
[2:54] uh that energy correlates with almost
[2:56] any metric of quality of life. So, if
[2:59] you have a couple hours to waste, I
[3:00] don't suggest going here unless you do,
[3:02] cuz it's really interesting website. You
[3:03] can make all these really interesting
[3:05] plots, uh correlations of data. So, you
[3:07] can plot energy use per person, pick
[3:09] your favorite metric, child mortality,
[3:11] you know, human uh UN human development
[3:13] index, average income per person. You
[3:15] don't have to get MATLAB out to do a
[3:16] regression fit, you know, these are
[3:17] pretty tightly correlated. So, this is
[3:19] good. You know, but what's the problem
[3:21] here? So, you know, the problem is that
[3:23] those correlations are nearly identical
[3:25] when you plot them against CO2 use per
[3:27] person. And this makes us face, you
[3:29] know, the the hard reality of climate
[3:31] change, which is that every single
[3:33] nation that's ever industrialized and
[3:35] made a better life for its citizens did
[3:37] so at the expense of the climate. There
[3:39] is no other way to do it presently than
[3:41] to dump carbon dioxide into the
[3:43] atmosphere by burning fossil fuels.
[3:46] So, you know, we think, you know, at the
[3:47] PSFC that fusion energy has a role to
[3:49] play in this problem. So, what is
[3:50] fusion? You know, most simply, you take
[3:52] two nuclei, you put them together, it
[3:54] releases energy.
[3:55] Uh it has immense promise. It's
[3:57] compelled this field of research for
[3:58] almost 60 years. So, you know, what what
[4:00] does that mean? Well, let's take one
[4:01] particular type of fusion. Take water,
[4:03] make deuterium, put a little bit of
[4:05] energy in, out comes, you know, 10 to
[4:07] infinity times the energy you put in.
[4:09] You know, this looks really attractive.
[4:11] And it is really attractive when you
[4:12] stop and and and look at it. So, what
[4:14] are some advantages of fusion over
[4:16] fossil fuels? No carbon, no socks or
[4:18] knocks, no particulate emissions.
[4:20] Inexhaustible fuel supply. Fuels are
[4:22] equally accessible around the globe. You
[4:24] know, there's no large-scale extraction
[4:25] or transport of fuels.
[4:28] Fusion looks pretty good compared to
[4:29] renewables in some ways. You know, high
[4:30] power density, efficient land use, high
[4:33] power density of the materials that
[4:35] you're using. You know, it's on when
[4:36] it's wanted, when it's needed. You can
[4:38] cite it where it's needed. It goes into
[4:40] established electricity grids without
[4:42] complicated energy storage technologies
[4:44] that don't yet exist. There advantages
[4:46] of fusion over fission energy. So,
[4:48] there's no physical possibility of a
[4:50] meltdown. You know, there's no
[4:51] long-lived nuclear waste that requires
[4:53] deep, long-term storage and
[4:55] repositories. And there's a very low
[4:57] proliferation risk in fusion. So,
[4:59] there's no need for uranium, plutonium,
[5:01] essentially bomb materials. You know,
[5:02] and clandestine uses of a fusion device
[5:05] to make those materials is very, very
[5:07] unlikely and easily detectable.
[5:10] Okay. So, you put all these advantages
[5:12] together, and there's a lot of people
[5:14] who think this is exciting. I'm going to
[5:15] get rich. I'm going to make my billions
[5:17] off of fusion energy, and I'm going to
[5:18] solve a lot of problems while I'm at it.
[5:21] So, this this compelling aspect of
[5:24] fusion, you know, it means there are a
[5:25] lot of different ways that people are
[5:27] trying to get there. Okay? And there's
[5:30] so many approaches that it can be
[5:32] extremely difficult to figure out which
[5:34] of these approaches, and this is only a
[5:35] few, are winners, which are long shots,
[5:38] and which are real losers. Okay? And I
[5:40] think this is true of, you know, your
[5:42] average layperson, you know, somebody
[5:44] who's read Time magazine recently, the
[5:46] investor, and to be quite frank, even
[5:48] fusion scientists when presented with a
[5:50] lot of these concepts might have trouble
[5:51] articulating, you know, what is the key
[5:53] point that I need to assess you know,
[5:56] what these technologies look like. Are
[5:59] they going to make fusion energy?
[6:01] You might say I'm not an expert, I'm not
[6:02] an expert in fusion science, you know,
[6:04] how can I judge these things? But if we
[6:06] were to turn the clock back, let's say
[6:08] 100 years, and look at a different
[6:09] complicated technology, the picture gets
[6:12] a little bit easier. Okay? So, the
[6:14] year's 1905, someone approaches you to
[6:15] invest in something. Okay? Here's the
[6:18] pitch. I've come to you, I've engineered
[6:21] this incredibly complex form of deluxe
[6:23] engineering. It's got bells and
[6:24] whistles, you know, engines, I can go
[6:26] fast, I've got flight attendants that
[6:28] can deliver you beverages at 40,000 ft.
[6:30] This looks really nice. You know, your
[6:32] your question to this person would be
[6:35] I see where you're coming from. Where's
[6:37] the wing?
[6:39] Right? You can't fly without a wing.
[6:41] It's the most complicated physical part
[6:42] of an aircraft. So, in the same way,
[6:44] somebody else might come to you. They
[6:46] might show you a picture and they say,
[6:47] "Look, we've worked out basic
[6:48] aerodynamics. Okay, we understand lift,
[6:51] we understand wing shear, here are some
[6:52] equations. More importantly, you know,
[6:55] we put a balsa glider together and we're
[6:56] pushing people off hills. Nobody's died
[6:58] yet. So, this looks good. Okay? You
[7:01] know, the key difference here, you
[7:02] should think about this is a physics
[7:03] problem.
[7:05] Okay? Going from something with the
[7:06] principles under understood, you know,
[7:08] basic demonstrations, an engineering
[7:10] problem. They're complicated, but
[7:11] there's some important qualitative and
[7:13] quantitative differences in tackling
[7:15] these problems.
[7:16] So, the purpose of this talk is to give
[7:19] you tools, simple tools that you can use
[7:21] to evaluate all these different
[7:23] approaches. Right? So, what are the
[7:24] right questions that you should ask when
[7:26] you're confronted with a type of fusion
[7:28] energy? Okay?
[7:29] And the secondary purpose is to use that
[7:31] information to motivate why at the PSFC
[7:34] we think there's a new pathway to fusion
[7:36] energy
[7:37] that is based in a technology called the
[7:39] tokamak that uses very high magnetic
[7:41] fields to confine a plasma and create
[7:44] the conditions required for fusion
[7:45] energy.
[7:47] So, the outline is sort of based on
[7:49] questions. So, we'll develop three rules
[7:52] for assessing fusion energy as answers
[7:54] to three questions. And then we'll go
[7:56] through and we'll take a sort of an
[7:57] abbreviated look
[7:59] at MIT's, you know, the PSFC's proposal
[8:02] for this pathway to net fusion energy.
[8:03] So, let's start here. What are the
[8:05] viable fusion fuels? How do they affect
[8:07] the approach?
[8:10] So, the first thing you have to
[8:11] understand about nuclei is they exist.
[8:13] There are protons and neutrons. They are
[8:15] held together by something called the
[8:16] strong nuclear force, which helps
[8:18] overcome the Coulomb repulsion between
[8:20] the charged particles. So, this
[8:22] these two forces keep things together in
[8:25] certain configurations that we call the
[8:26] nuclei.
[8:27] And what you have to understand about
[8:29] nuclear physics is quite simple. Some
[8:31] nuclei like to be together. They require
[8:34] less energy to hold them together. Other
[8:36] nuclei do not like to be together, and
[8:39] you actually have to put more energy in
[8:40] to hold them together. It's something
[8:41] called binding energy. So, something
[8:43] like hydrogen, which isn't bound at all,
[8:45] and something like helium, you know,
[8:47] some of the things here that are weakly
[8:48] bound require more energy to hold them
[8:50] together. Iron, in particular, is a
[8:52] nucleus that likes to be together.
[8:54] Doesn't require much energy.
[8:56] Changing the configuration of the nuclei
[9:00] can potentially make some of that
[9:01] binding energy, which is no longer
[9:03] needed to keep the nuclei together,
[9:05] available to do things with. So, many
[9:07] people might be familiar with fission.
[9:09] Take a heavy thing like uranium, you
[9:11] know, fission it down into something
[9:13] smaller. These things, these products,
[9:15] are happier to be together. They don't
[9:17] require as much of this binding energy
[9:18] binding energy to hold together. You can
[9:20] use that energy to go do something with.
[9:22] You know, fusion is the opposite
[9:24] process. Take small things that are
[9:25] weakly bound, push them together,
[9:28] release binding energy, now you can do
[9:30] something with it. Okay? It's the basics
[9:32] of fusion.
[9:33] So, what do I need? What things do I
[9:36] need to know to evaluate what I should
[9:38] try to fuse? Okay? So, there's basically
[9:40] two physical quantities. There's the
[9:41] reaction energetics, i.e., the input
[9:44] energy, the output energy,
[9:46] and there's the reaction probability.
[9:47] So, let's start with the input energy.
[9:49] Okay? When I take two charged particles,
[9:51] I try to push them together, they want
[9:52] to repulse with the Coulomb force. So, I
[9:54] have to put some amount of energy in
[9:57] to overcome this barrier. You can
[9:59] calculate what that energy is with some
[10:00] very simple physics, okay? So, I have to
[10:03] have enough energy to push these things
[10:05] together. That's the input energy.
[10:07] The output energy is very important. So,
[10:09] whatever output energy I get from my
[10:11] fusion, from binding energy, has to be
[10:13] large, you know, the larger the better
[10:14] so that I have enough energy to convert
[10:16] into electricity.
[10:18] The second quantity then is probability.
[10:20] So, you know, the technical term we use
[10:22] is a cross section, but what you should
[10:24] think of is probability that when I
[10:26] bring these two things together, they'll
[10:27] fuse.
[10:28] So, it has to be high, okay? The cross
[10:30] section is not simple, you know, if I
[10:32] use this simple formula and I treat
[10:33] these things like billiard balls when I
[10:34] bring them together, you know, it's like
[10:36] threshold condition, either they don't
[10:37] fuse or they do fuse. Turns out things
[10:39] are a little more complicated, there's
[10:40] quantum and nuclear physics involved,
[10:42] and the cross section has some
[10:43] complicated shape, okay? There's a whole
[10:45] scientific discipline involved in simply
[10:48] measuring what those shapes are, but
[10:50] they turn out to be very important.
[10:52] So, our ideal fusion fuel has three
[10:54] things. Low input energy
[10:56] required to induce a reaction. That's
[10:58] important because it's technologically
[11:00] easier to achieve, and it's economic
[11:03] because it requires less input energy,
[11:05] right?
[11:06] So, a high probability of fusion, you
[11:08] know, higher the better as far as we're
[11:09] concerned, and a high output energy for
[11:11] converting to electricity. Higher the
[11:13] better.
[11:14] So, what are our options, okay? So, our
[11:16] options in fusion
[11:17] are the 3,000 nuclides that exist in the
[11:20] known universe. So, we can look at any
[11:22] of these things and we can try to
[11:23] evaluate fusion.
[11:24] Does it work? Is it useful? So, what
[11:26] we'll do is we'll start in this bottom
[11:27] corner. So, I'll plot hydrogen, you
[11:29] know, we'll start moving towards heavier
[11:31] elements.
[11:32] Same thing over here. So, we have to
[11:33] take two things, we have to fuse them
[11:35] together. So, what does this look like?
[11:38] If we were to look at the heavier
[11:39] elements, things carbons, berylliums,
[11:42] you know, lithiums, you know, anything
[11:44] basically heavier than sort of where
[11:46] this line is, the Coulomb barrier is
[11:48] simply too high. The energy required to
[11:50] fuse those things together is not really
[11:52] feasible to achieve in the right
[11:53] conditions. So, we simply leave those
[11:55] out. Not usable.
[11:57] Next, what we'll do is look at the
[11:58] probability that any of these two will
[12:01] actually fuse when I bring them
[12:02] together. And, you know, don't pay
[12:04] attention to the little numbers,
[12:06] but what you can basically see is there
[12:07] are some probabilities that are so
[12:08] small, we don't even bother with them.
[12:10] They're just not useful from the get-go,
[12:12] okay?
[12:14] You know, and then when I look at the
[12:15] energy output, and I look at this
[12:17] probability that they'll fuse, most of
[12:19] the rest of these are sort of marginal,
[12:21] all right? They're kind of on the
[12:22] borderline. Maybe you could use them.
[12:25] It's pretty clear that they're going to
[12:26] prevent you doing anything really
[12:29] practical in the near term with known
[12:31] technology.
[12:32] So, that leaves very few options. In
[12:35] fact, it leaves three, okay? So, let's
[12:37] take a look at what those three
[12:38] reactions are, and just for the sake of
[12:40] the argument, because this is a popular
[12:42] fuel that some fusion energy attempts
[12:44] are trying to use, we'll include a
[12:45] fourth one, one of these marginal fuels,
[12:48] proton and boron, okay? So, you can see
[12:50] there's input particles, you know,
[12:52] deuterium and tritium, these are types
[12:54] of hydrogen. There are output particles.
[12:57] There's energy gain, and then there's
[12:58] the probability that something will fuse
[13:01] at a specific energy that's required to
[13:03] bring them together, okay?
[13:06] Now, there's a slight subtlety in this,
[13:09] okay? And the subtlety is a thing called
[13:11] Coulomb scattering. You should think of
[13:12] this sort of as the bogeyman of fusion,
[13:14] all right?
[13:16] If you want to fuse particles, okay,
[13:17] that's great. You have some probability
[13:19] of doing so as a function of energy, but
[13:21] it turns out that the probability that
[13:23] those particles when you bring them
[13:24] together will simply bounce off each
[13:25] other is always enormously larger than
[13:28] the fusion probability. That turns out
[13:30] to be a key point in any approach to
[13:33] fusion energy.
[13:35] So, you know, what what might be kind of
[13:36] an analogy here? If I have no way to
[13:39] confine these particles, they're simply
[13:41] going to scatter and they're lost,
[13:42] right? I think of a billiard ball table
[13:44] with no bumpers around the edge, you
[13:46] know, you throw a bunch of billiard
[13:47] balls together, they might knock into
[13:49] each other, they're never going to go in
[13:50] the pockets cuz the probability is low,
[13:51] they just fall off the table. It's not a
[13:53] particularly useful system.
[13:55] You know, but if you were able to come
[13:56] up with some ideal confinement scheme,
[13:58] who cares if these particles scatter?
[14:00] Let them scatter all day if you want.
[14:01] Eventually, they're going to fuse, and
[14:03] if you get enough fusions, life is good,
[14:05] you've got energy.
[14:07] Now, when we confine these particles at
[14:09] the energies required to get to the
[14:11] fusion cross section,
[14:13] that creates the conditions of something
[14:14] we call a plasma. So, a plasma is an
[14:16] ionized gas, it's a good way to think of
[14:18] it, electrons and ions, kind of think of
[14:20] them as two fluids that move together.
[14:23] You know, the behavior of a plasma is
[14:24] not simple, it's not described by the
[14:26] simple motion of individual particles,
[14:28] but by their collective behavior, and
[14:30] the energy of that system is best
[14:32] described as a temperature, right? So,
[14:34] if you have a balloon that's got hot air
[14:36] in it, we don't talk about the speeds of
[14:37] the gas molecules, we talk about its
[14:39] temperature cuz it defines the the sort
[14:41] of collective behavior. So, from here on
[14:43] out, what we're going to do is talk
[14:44] about temperature rather than energy,
[14:46] but this effectively you should think of
[14:47] them as the same thing.
[14:49] So, let's return to our fuels, okay? So,
[14:52] we've now changed the probability that
[14:54] something will fuse to how much fusion
[14:56] happens per unit volume per second, and
[14:58] we'll actually rank these fuels, okay?
[15:01] And so, the number one easiest fuel to
[15:03] use in fusion is deuterium and tritium.
[15:06] Why is that? The The reason
[15:09] there's sort of a primary condition
[15:10] here. The primary condition is that 15
[15:13] keV, or about 175 million degrees,
[15:16] Kelvin, Celsius, Fahrenheit, doesn't
[15:18] matter at that scale,
[15:20] um, is actually low temperature, believe
[15:22] it or not, in the fusion world.
[15:23] As you go to, you know, increasingly
[15:25] difficult fuels, your temperature goes
[15:28] up. You know, by the time you're doing
[15:29] proton boron 11 fusion, you better have
[15:31] a plasma which is close to 2 billion
[15:33] degrees.
[15:34] This is a technologically very, very
[15:37] challenging. Uh, the secondary condition
[15:40] here
[15:41] is you want a lot of energy out,
[15:43] and you want a lot of fusions to happen
[15:45] in whatever volume you've confined
[15:47] things in. So, you can see here, the DT
[15:49] fusion, sort of in any way you rank it,
[15:51] is easily the sort of the best way to
[15:54] go, and it's usually the fuel of choice
[15:55] for most fusion concepts, but other
[15:58] concepts have picked different different
[15:59] fuels.
[16:01] What do you do with those fuels? You
[16:03] know, how is this fusion energy
[16:04] expressed,
[16:05] and how do you turn it into electricity?
[16:07] So,
[16:08] deuterium and tritium, we bring them
[16:09] together, they fuse. The energy that's
[16:11] released, so that binding energy that's
[16:13] available, comes out as the kinetic
[16:15] energy, right? The speed of these
[16:17] particles, okay? So, if I have a fuel
[16:19] cycle that generates neutrons, I have a
[16:21] nice option.
[16:23] I can take those neutrons, I can let
[16:25] them rattle around in a material and
[16:27] turn their kinetic energy into heat. I
[16:29] can hook up a,
[16:30] you know, conversion, heat conversion. I
[16:32] can drive make steam drive a turbine and
[16:34] make electricity.
[16:36] There's an additional benefit of having
[16:37] a neutron for this particular fuel
[16:39] cycle, which is if I'm clever,
[16:41] I can actually generate my own fuel. So,
[16:43] I can make more fuel, put it back into
[16:45] my reaction. It's a nice little trick.
[16:48] Now, I have these alpha particles, these
[16:50] helium nuclei, these are confined in
[16:52] some way, you know, and typically these
[16:54] come out at much higher temperature or
[16:56] energy than the initial fuel, and so as
[16:58] these collide with those other
[17:00] particles, they give their energy back.
[17:01] They keep the reaction hot and
[17:04] energetic. You know, and eventually
[17:05] these particles
[17:07] are exhausted from the system, but along
[17:09] the way they collide with electrons,
[17:10] they radiate photons,
[17:12] and they they lose energy. But, you
[17:14] know, you might feasibly imagine taking
[17:16] some technology
[17:18] and capturing that x-ray heat and
[17:20] turning that into electricity.
[17:23] Now, of course, if I don't have a
[17:24] neutron, that changes the way I can
[17:26] convert fusion energy into electricity.
[17:29] You know, again, I still have charged
[17:30] particles that heat, you know, whatever
[17:32] fuel I still have confined,
[17:34] but now I now have x-rays, you know? So,
[17:36] again, I can I can try to do this x-ray
[17:38] capture into heat and drive a turbine,
[17:40] or
[17:42] I could come up with some sort of
[17:43] technology called direct energy
[17:45] conversion, which takes the kinetic
[17:47] energy of charged particles,
[17:49] and through some slightly speculative
[17:52] technology at this point, creates
[17:54] electricity. Okay, so those are my
[17:55] options. So, let's answer our first
[17:57] question, which is what are viable
[17:59] fusion fuels? How do they affect the
[18:01] approach?
[18:02] Rule number one is fuel choice
[18:03] fundamentally
[18:07] sets the difficulty of any approach to
[18:09] fusion energy. So,
[18:10] DD and D helium 3 are increasingly
[18:12] difficult, and proton boron 11 is
[18:15] possibly infeasible. Other fuels are
[18:17] simply not viable. So, when you
[18:19] encounter fusion concepts, you read
[18:20] about them in Time, the newspaper, you
[18:22] know, etc., you know, ask questions.
[18:24] What fusion fuel are they using? You
[18:26] know, do they acknowledge the
[18:27] difficulties of each type of fuel?
[18:30] How do they propose conversion to
[18:31] electricity? You know, how mature and
[18:33] demonstrated is that technology if
[18:35] they're making reactor claims today?
[18:37] Okay?
[18:39] Let's move on to question two. So, what
[18:41] are the physical conditions required to
[18:44] achieve net fusion energy?
[18:46] So, most people in this room, in fact,
[18:48] no one in this room, because we've never
[18:49] burned a plasma,
[18:51] have experience with burning plasmas,
[18:52] but most people probably have experience
[18:54] with burning wood. You've tried to make
[18:56] a fire at home, camping, etc.
[18:59] And it turns out the conditions for
[19:00] burning a plasma are analogous to the
[19:02] conditions for burning wood. So, you
[19:03] need enough wood, but not too much.
[19:06] You need to get the temperature of the
[19:07] fuel up. So, if you've ever tried to
[19:09] light a wet campfire, you know that the
[19:11] heat you're putting in goes into
[19:12] creating steam, you're heating up the
[19:13] water, not the wood. And of course, once
[19:16] you have a fire going, you have to keep
[19:18] the heat around. So, specific
[19:20] arrangements of logs do a good job of
[19:22] confining the heat and continually
[19:24] generating,
[19:26] you know, driving the reaction forward.
[19:28] So, in the plasma world, it's the same
[19:30] thing.
[19:31] I need the right density, not too much,
[19:33] not too little. I need the temperature
[19:35] to be right where I want it so that I
[19:37] can get high fusion reactivity, and I
[19:39] have to come up with some scheme, just
[19:41] like in the fire, that confines the heat
[19:44] into my reaction that continually drives
[19:47] things forward, okay? So, density,
[19:50] temperature, and energy confinement,
[19:52] which we call it energy confinement
[19:54] time. How long does it take heat to leak
[19:56] out of the system?
[19:57] These have been known for a long time.
[19:59] So, 1955, J. D. Lawson derived this
[20:01] thing we call the Lawson criterion,
[20:03] which uses these three simple things to
[20:06] set the conditions required for a
[20:07] burning plasma.
[20:10] There's a nice way to visualize this.
[20:12] So, if we were to plot density times
[20:14] confinement time, often called the
[20:15] Lawson parameter or Lawson criterion,
[20:18] versus the temperature of the plasma, we
[20:20] get this space. Okay? This space spans
[20:24] many decades in temperature and many
[20:25] decades in the Lawson parameter.
[20:28] And then we can draw these contours. So,
[20:30] what is capital Q? Capital Q is an
[20:32] important parameter in fusion. It's
[20:33] fusion energy output divided by the
[20:35] energy input. If you want to make net
[20:38] energy and put watts in the grid, you
[20:40] clearly want a high capital Q,
[20:42] preferably greater than one. Otherwise,
[20:43] you're consuming energy. Okay?
[20:47] Now,
[20:49] 60 years of fusion research is basically
[20:51] about picking a concept,
[20:53] probably starting down here or off the
[20:55] plot, and trying to move into this
[20:58] region.
[20:59] That's the magic space to be in for
[21:01] fusion energy. But,
[21:03] it turns out this is enormously
[21:05] difficult in practice. Okay? So, think
[21:07] about this. You are moving many orders
[21:09] of magnitude in absolute parameters,
[21:12] which is fundamentally changing your
[21:14] physical system. Okay? So, often, there
[21:17] are unknown unknowns as you move through
[21:20] this space. So, we call them plasma
[21:21] instabilities, for example. You don't
[21:23] need to know what that is. You simply
[21:25] know that these things are often lurking
[21:27] out in this space waiting to destroy
[21:28] your dreams of taking a concept and
[21:31] making fusion energy. So, many concepts
[21:33] which start off encounter one of the
[21:35] myriad of instabilities that you can go
[21:37] out and look. Sometimes those
[21:39] instabilities can be overcome with great
[21:40] time and effort and money. Sometimes
[21:42] they simply can't, and a concept is
[21:44] abandoned.
[21:47] So, what does this mean for fusion
[21:48] energy concepts? All right. Let's take
[21:50] this space. Let's divide it into three
[21:51] regions. Let's make a comparison.
[21:54] We've all been to a theme park, roller
[21:56] coasters and big rides. Okay? So,
[21:59] sorry, you are not able to ride. You're
[22:01] not tall enough.
[22:03] Here, you must be accompanied by an
[22:04] adult. You know, and in green, you can
[22:06] enjoy your ride. All right. You're tall
[22:07] enough to ride. You're ready to go. So,
[22:09] fusion energy, the ride.
[22:11] So,
[22:13] if you are in this red space, your
[22:15] fundamental physics is simply not yet
[22:17] viable if you want to talk about energy
[22:19] production.
[22:20] Reactor engineering at this point is
[22:23] putting the cart in front of the horse.
[22:25] You're simply not ready to think about
[22:27] what a reactor looks like cuz you
[22:28] haven't even gotten the physics in the
[22:30] right region. You know, and certainly
[22:32] grandiose reactor claims that you're
[22:33] going to come up with watts in the grid
[22:35] in 3 years are exploitative of people
[22:37] who don't know enough to understand
[22:39] even, you know, the physics behind what
[22:41] you're trying to do. Okay?
[22:43] If you're approaching this yellow space
[22:45] or in this yellow space, your physics is
[22:47] nearing demonstration. You are now
[22:49] demonstrating quantitatively that you're
[22:51] in the right spot for fusion energy.
[22:53] Thinking about reactor engineering or
[22:55] doing it seems justified at this point,
[22:57] and reactor claims are reasonable.
[22:59] Right? We sort of know the physics. We
[23:00] kind of know what our particles are
[23:01] coming out. We sort of know what
[23:03] technologies are needed.
[23:05] Seems reasonable. If you're in green,
[23:07] life is good. You know, physics is
[23:09] unquestionably demonstrated. Your
[23:10] reactor engineering should be mature,
[23:12] and your reactor might even be putting
[23:13] watts on the grid. Okay? So, so that's
[23:15] where we're at.
[23:17] So, let's ask Q2.
[23:18] What are the physical conditions
[23:19] required to achieve net fusion energy?
[23:21] Well, it's density, temperature, and
[23:23] confinement time. So, the proximity to
[23:26] burning plasma conditions, right? I.e.,
[23:28] those regions in the top right corner is
[23:30] the ultimate arbiter of the viability of
[23:32] any fusion energy approach. Okay? So,
[23:35] temperature, if your temperature
[23:37] and your Lawson parameter, your density
[23:39] times confinement time, give a Q greater
[23:41] than about 0.1,
[23:43] you're sort of there. Right? You're
[23:44] ready for fusion energy. If your
[23:46] temperature and your density and
[23:47] confinement time give you a a net
[23:49] capital Q again of about 0.1 or less,
[23:53] you're physics experiment. You have a
[23:54] long uncertain pathway in front of you
[23:57] to get where you need to be. So,
[23:59] questions you should ask when you
[24:00] encounter a fusion concept. You know,
[24:02] what is the plasma pressure?
[24:03] What is the ion temperature? What's your
[24:05] confinement time? You know, what capital
[24:07] Q values, i.e., energy gain, are
[24:09] achieved on your present machine? You
[24:11] know, how many orders of magnitude do
[24:13] you have to extrapolate in this space to
[24:15] get to a reactor relevant condition on
[24:18] your current performance? What is, you
[24:19] know, is your plasma stable? Do you have
[24:21] problems with turbulence causing you to
[24:23] leak energy out of your confinement
[24:25] scheme? Okay?
[24:28] So, let's turn to the third and final
[24:29] question. So, what fusion energy
[24:31] approaches exist?
[24:33] How should we evaluate them?
[24:36] It turns out, again, that there are an
[24:39] enormous number of ways to confine
[24:41] particles and try to make the conditions
[24:43] for fusion energy. Okay? This is a nice
[24:46] hand sketch plot by Simon Woodruff from
[24:48] about 15 years ago that I like,
[24:50] and it certainly includes does not
[24:52] include all of the potential ways of
[24:54] doing it, but it includes a lot. And it
[24:56] roughly divides things up into, you
[24:57] know, magnetic fusion and inertial
[24:59] fusion, open field lines, and other
[25:02] things. But, what we're going to try to
[25:03] do is go through some of the key
[25:04] concepts, particularly ones that maybe
[25:07] you've read about recently, and try to
[25:09] assess, you know, how close are they?
[25:11] Right?
[25:13] So, unfortunately, we have to start
[25:14] here, cold fusion,
[25:16] alias low energy nuclear reactions or
[25:18] LENR, now that this is sort of
[25:20] scientifically unpalatable. Okay? So,
[25:23] the confinement basis
[25:25] unknown.
[25:26] Active experiments, there's a
[25:27] surprisingly large number of them still
[25:29] going.
[25:30] You know, the key takeaway here is if
[25:32] cold fusion is too good to be true, it
[25:34] probably is. Right? So, cold fusion is
[25:37] basically a, you know, purports to use
[25:39] some physical process to create fusion
[25:40] energy conditions. You know, it was
[25:42] first discovered in the the late 1980s
[25:44] by Pons and Fleischmann famously. And
[25:46] it's important to point out that the
[25:48] proposed process cannot be rectified
[25:50] with any known physical model of the
[25:52] universe.
[25:53] Not to say maybe there isn't one out
[25:55] there,
[25:56] but it certainly can't currently be fit
[25:58] into the way we understand the universe.
[26:01] This has been rapidly and continually
[26:02] debunked over the years. There's zero
[26:04] independent validation by critics, and a
[26:06] lot of the initial debunking work was
[26:08] actually done right here at MIT by a
[26:10] couple of the professors in the the
[26:11] plasma science and fusion center. So,
[26:13] it's a great picture of the Boston
[26:14] Herald saying, you know, the bombshell
[26:16] from MIT, you know, destroyed cold
[26:18] fusion. And in fact, cold fusion is such
[26:20] an interesting case that it is in self
[26:23] itself studied by other sciences. So,
[26:25] like science and technology studies take
[26:27] a look at this and call it a
[26:28] pathological science, something people
[26:30] keep doing even though it's been proven
[26:32] to be false.
[26:33] And so, how are they doing? Well,
[26:35] without being flippant,
[26:37] you can't actually put a zero on a
[26:39] log-log plot. Right? So, we're simply
[26:42] going to say that cold fusion, you know,
[26:43] not a number. Doesn't exist in this
[26:45] space that we're using to evaluate
[26:47] fusion.
[26:48] Okay, let's try something different. So,
[26:50] gravity is another force. It's a force
[26:51] we could use.
[26:53] And it's pretty clear that the universe
[26:54] has figured out how to do fusion. So,
[26:57] if you were to take the sun and look at
[27:00] it, you know, what the sun is is a ball
[27:01] of plasma. It is fusing hydrogen to
[27:03] helium to lithium to beryllium and so on
[27:05] up.
[27:06] And that, you know, the conditions of
[27:08] fusion there, this burning plasma wants
[27:10] to expand, and gravity wants to
[27:12] contract. And so, you eventually you end
[27:14] up with this transient equilibrium,
[27:16] transient meaning billions of years in
[27:17] this case. But, you know, the sun is
[27:20] able to produce a massive amount of
[27:21] power by fusing things together with
[27:23] gravity.
[27:24] Important to point out, it's a tiny
[27:26] power density, you know, about a third
[27:28] of a watt per cubic meter. That's sort
[27:29] of your average compost pile. That's the
[27:31] heat it generates, volumetric heat. But,
[27:34] of course, it does it over an enormous
[27:36] volume. So, you end up getting out quite
[27:38] a bit of power. Okay? So, this is a
[27:39] balance of plasma pressure, think of a
[27:42] balloon expanding, with gravity pushing
[27:44] back. Okay?
[27:46] You know, the technological complexity
[27:48] of replicating this on or near Earth at
[27:49] some point is pretty small. So, this is
[27:52] not viable. But, if we were to evaluate
[27:54] the sun
[27:55] on our space here, you know, the sun is
[27:56] clearly
[27:57] well up this curve. It's making net
[27:59] energy.
[28:02] Okay, let's try a different approach.
[28:03] You know, what if we did this with
[28:04] bombs? Right? Nuclear bombs. So, what
[28:09] you could do, what is what is a hydrogen
[28:10] bomb? It's basically a primary, a
[28:12] fission bomb, that uses uranium or
[28:14] plutonium.
[28:15] It implodes, generates X-rays. Those
[28:18] X-rays then crush a fusion secondary and
[28:21] create the conditions by crushing the
[28:23] fuel down to generate net energy gain.
[28:26] So, it's important to point out, you
[28:28] know, to date, this is the only
[28:29] successful net fusion energy gain on
[28:31] Earth.
[28:33] Only time we've done it on Earth. Now,
[28:34] it's in a controlled fashion, and this
[28:36] is probably not a great power source.
[28:38] Right? There goes the power plant. So,
[28:41] even though bombs, you know, clearly
[28:42] meet our criterion for a fusion reactor,
[28:45] not particularly useful.
[28:47] Okay, but this idea of crushing fuel is
[28:49] kind of interesting. So, let's keep on
[28:50] that. So, what if we use some other
[28:52] method to try to crush fuel?
[28:55] So, inertial confinement fusion
[28:58] is what it's called. There's various
[28:59] ways that you can do it.
[29:01] In this case, you might use lasers. So,
[29:03] the National Ignition Facility is an
[29:04] incredibly impressive experiment that
[29:06] uses lasers to drive this implosion
[29:09] process. You know, and eventually those
[29:11] lasers aren't on holding the fuel in
[29:13] this state. You know, that fuel
[29:14] eventually expands, you know, and blows
[29:16] the the capsule apart. So, you know,
[29:19] this is an impressive experiment, and
[29:21] they've demonstrated some really nice
[29:22] physics results. Okay? It's hard to
[29:24] compare inertial confinement fusion to
[29:27] you know, to any other type of fusion,
[29:29] but it is a little different. But, we
[29:31] can make an estimate for what their
[29:32] capital Q is. It's about 0.1, which is
[29:36] pretty good. They're pretty close to
[29:37] actually getting above this, you know,
[29:40] capital Q of
[29:41] factor. But, they have some
[29:42] difficulties. So,
[29:44] one of the key things, you know,
[29:46] maintenance of such a machine. So,
[29:48] significant portions of the machine in
[29:50] this approach are destroyed every time
[29:52] you implode a capsule. So, there's no
[29:53] way around that at this point. You know,
[29:55] the number of times that you implode a
[29:57] capsule is very important. You know, and
[29:59] at their current rate, you need to speed
[30:01] up the number of times they do that from
[30:02] about once a day to about once a second.
[30:04] That's a factor of 100,000. It's very
[30:06] technologically difficult to imagine how
[30:08] to do that. You know, and their
[30:09] efficiency is is difficult. So, you take
[30:12] a certain amount of power
[30:13] and you try to transmit that into your
[30:15] target, and the efficiency is really
[30:17] low. So, it seems difficult to imagine
[30:19] in the near term that these are a
[30:21] compelling way to get to fusion energy.
[30:24] Okay, so pushing things together doesn't
[30:25] seem to work that well.
[30:27] What if we take a different force? Let's
[30:29] take the electric force, okay? So, we
[30:32] can take an electric field
[30:34] and we can generate some deuterium and
[30:36] tritium ions and slam them into a target
[30:38] and try to get to the conditions of
[30:39] fusion energy. So, push particles
[30:41] together and hope they fuse.
[30:43] Now,
[30:44] let's remember back to the bogeyman of
[30:45] fusion, Coulomb scattering, right? The
[30:47] Coulomb cross section is far too large
[30:50] for this approach to ever work. Too many
[30:52] scatters,
[30:54] too few fusions, and those particles are
[30:55] not being kept around long enough to get
[30:58] to the conditions that you want.
[31:00] Now, this is fusion, I'm afraid. This is
[31:02] a field full of optimists, and this
[31:04] didn't stop people from trying. So,
[31:06] there's various experiments, Migma 4 was
[31:08] one,
[31:09] and they were able to create, not
[31:11] surprisingly, you know, very high energy
[31:13] particles, but very very poor density
[31:15] and confinement time. So, they don't
[31:16] even appear on this plot.
[31:19] Okay, well, if a linear particle
[31:21] accelerator doesn't work, what if we
[31:22] make a three-dimensional spherical
[31:24] particle accelerator? Maybe that'll
[31:26] solve the problem. There are different
[31:27] ways that you can go about doing this,
[31:29] but you know, effectively what you're
[31:31] trying to do is use some configuration
[31:32] to generate an electrostatic potential
[31:35] that creates a well in the middle of a
[31:36] sphere that ions will fall into
[31:39] and rattle around together and fuse. You
[31:42] know, again, it turns out Coulomb
[31:44] collisions are a problem. Losing
[31:47] particles out of the spherical well are
[31:49] a problem. You know, conduction losses
[31:51] are a problem. Radiating energy away
[31:52] through bremsstrahlung photon emission
[31:54] turn out to be problematic, okay?
[31:57] So, we can try to put a few of these
[31:58] experiments on the plot. You know,
[32:00] they're somewhere here. They don't quite
[32:01] work out. You know, it is important to
[32:03] point out that these are simple devices,
[32:05] but they're great for people just
[32:06] curious about fusion. Students, you
[32:08] know, people at universities, physical
[32:11] labs. So, they're very interesting. So,
[32:12] typically, if you read one of those
[32:14] articles, which is, you know, teenager
[32:15] produces fusion reaction at home,
[32:18] it's typically done with something like
[32:19] a fusor, which is a fantastic way to get
[32:21] interested in plasma physics and
[32:23] technology. Turns out to be a poor way
[32:25] to make energy.
[32:28] All right, so we're running out of
[32:29] forces, unfortunately. So, let's try
[32:31] magnetic forces. So, let's try magnetic
[32:33] fields. So, something that's important
[32:35] to understand about ions in a magnetic
[32:38] field is that if you have field lines,
[32:41] those particles will spiral around the
[32:44] field lines in effectively two
[32:45] dimensions. So, they are confined to the
[32:47] magnetic field line, but they are not
[32:49] confined along the magnetic field line,
[32:52] okay?
[32:53] So, you can imagine different
[32:54] configurations of field lines that try
[32:56] to keep the particles around long
[32:58] enough. So, we'll do that. So, one of
[32:59] the earliest was something known as a
[33:01] mirror. So, you have these coils,
[33:03] and the magnetic field lines are in
[33:05] green. You know, and the sort of idea
[33:07] here was that plasma particles, which
[33:09] are spiraling around, will sort of
[33:10] approach the neck, they reach a
[33:12] particular condition, and they will
[33:14] reflect. You know, and the idea was that
[33:16] they would just sort of bounce back and
[33:17] forth. You know, unfortunately, you
[33:19] know, nature wasn't kind. Turns out that
[33:21] didn't work very well. You know,
[33:22] basically, these mirrors just
[33:23] hemorrhaged leakage out of the ends. And
[33:26] there were all sorts of incredibly
[33:28] complicated and clever people who spent
[33:30] decades trying to plug the ends, you
[33:32] know, magnet systems and other things,
[33:34] but it turned out it just didn't work,
[33:36] okay?
[33:37] So,
[33:38] unfortunately, you know, our conclusion
[33:39] for this particular configuration was
[33:41] you might be able to do fusion, but in
[33:43] order to deal with the end losses, you
[33:44] make your machine kilometers long. That
[33:46] doesn't seem particularly feasible.
[33:49] So, we can go and look at our space, you
[33:51] know, density and confinement time, ion
[33:53] temperature, and we can put all those
[33:55] mirrors on it. And it's important, you
[33:57] may have heard of Lockheed Martin.
[33:58] Lockheed Martin is currently
[34:00] actively working in the fusion space.
[34:02] They've come out with a concept. And
[34:04] Lockheed Martin's concept doesn't
[34:05] actually fit on this space. So, they
[34:08] continue to be, you know, four, five,
[34:10] six orders of magnitude away from the
[34:12] area that needs they need to be in to be
[34:15] energy relevant.
[34:17] Okay, so crimping the ends of our
[34:18] magnetic field didn't work.
[34:20] What if we try something different? So,
[34:22] what if we try squeezing magnetic
[34:24] fields? You know, so does that work? So,
[34:26] there's different arrangements of
[34:27] magnetic fields that you can do,
[34:29] and the field has pretty much tried all
[34:31] of them,
[34:32] but there were a lot of problems. So,
[34:33] there were problems with instabilities,
[34:35] you know, these physical consequences
[34:37] that we didn't expect that prevented you
[34:39] from actually climbing up this curve.
[34:42] There were other issues with plasma
[34:43] cooling in the same way before you got
[34:45] to net energy conditions. You know, so
[34:47] again, if we were to look on our space,
[34:50] temperature, Lawson parameter, you know,
[34:52] there's a few points on here, including
[34:54] a company which is in Canada called
[34:55] General Fusion, which is trying to do
[34:57] something called magnetic target fusion,
[35:00] and they're sort of down in this region.
[35:02] So, pretty interesting physics that's
[35:04] going on in these devices, and they
[35:06] might be useful as a as a high power
[35:07] x-ray at present or a neutron source,
[35:09] maybe a particle accelerator. You know,
[35:11] but at present, they're very far away
[35:13] from the conditions necessary for fusion
[35:15] energy. And they have a long way to go
[35:17] to climb up this path.
[35:20] Okay, so the mirrors didn't work out too
[35:22] well. So, what if we just used a lot of
[35:23] them?
[35:25] Right? That's the sort of the, you know,
[35:27] shotgun approach. So, take a mirror,
[35:29] didn't work, add another one on, another
[35:31] one on, another one on, another one on,
[35:33] ad infinitum, until you wrap them around
[35:35] and you get a torus. So, there were a
[35:37] number of programs in the '70s and '80s
[35:39] that tried to do this. By far, one of
[35:40] the best fusion names of all time, the
[35:42] Elmo Bumpy Torus.
[35:44] But it turned out that symmetry
[35:46] was really important, and by losing that
[35:48] symmetry, by
[35:49] going around, you had a lot of
[35:51] instabilities that were added to the
[35:53] plasma, limited temperatures, and poor
[35:55] confinement. So, interesting plasma
[35:57] physics, but again, try to put Elmo on
[35:59] our space, and it doesn't quite fit.
[36:02] You know, but we did learn something,
[36:03] which is that symmetries turn out to be
[36:05] important. You have to understand
[36:06] symmetry of the plasma.
[36:09] Now,
[36:09] we can do something different with the
[36:10] plasma. We can try to twist the magnetic
[36:13] field lines around. So, there's a number
[36:15] of configurations called the field
[36:17] reversed, FR FRC, spheromax, and other
[36:20] things that basically use the
[36:21] self-organization of a plasma to create
[36:24] the condition of a twist in the magnetic
[36:26] field. There's a number of active
[36:27] experiments out there. Tri Alpha Energy
[36:29] is one. It's a company that's gotten a
[36:31] lot of press. They're doing some really
[36:32] interesting things with their
[36:33] experiment. You can kind of see it over
[36:36] here. You know, the the field is
[36:38] toroidal, so that seemed to, you know,
[36:40] work out in some configurations as long
[36:42] as you preserve symmetry. And what if we
[36:44] use the plasma to create this twist,
[36:46] okay?
[36:47] Unfortunately, these devices are fairly
[36:49] transient, they have a lot of short time
[36:51] scale issues.
[36:52] They've been studied widely over a long
[36:55] period of time, and no one's been able
[36:56] to make particularly much progress. So,
[36:58] again, very interesting plasma physics,
[37:00] but if we were to go back to our space,
[37:02] and we look at the many types of FRCs,
[37:04] spheromax, and even Tri Alpha, they are
[37:06] many orders of magnitude
[37:08] away from where they need to be.
[37:10] Okay, but we did learn an interesting
[37:11] lesson when we actually did some physics
[37:13] with these things, which is that this
[37:15] helical confinement, so you have
[37:16] twisting magnetic field lines, turned
[37:18] out to be a really interesting way to
[37:20] tame certain types of stabilities.
[37:24] Okay, so instead of letting the plasma
[37:26] organize its own field, let's use coils.
[37:28] So, there's a type of configuration
[37:31] known as the stellarator, right? The
[37:33] stellarator uses these incredibly
[37:34] complicated magnetic field coils to
[37:36] create this sort of plasma shape that
[37:39] you see twisting around in yellow.
[37:41] This was one of the original fusion
[37:43] concepts. It's an active area of
[37:44] research. You know, one of the sort of
[37:46] most exciting fusion machines out there
[37:48] is Wendelstein 7-X in Germany. So, you
[37:50] can see a person here surrounded by
[37:52] these immensely complicated magnets.
[37:54] And they've actually done pretty well,
[37:56] right? This is the first sort of
[37:58] configuration of devices that we've seen
[38:00] sort of march out of this bottom left
[38:02] corner and make some progress towards
[38:05] the right conditions. So, they're not
[38:06] quite there yet. They've got a little
[38:08] bit of ways to go. There's some complex
[38:10] engineering issues with these devices,
[38:12] but, you know, they're not doing too
[38:13] bad. They'll be really interesting to
[38:15] watch over the next 10 years.
[38:17] So, the final configuration that we can
[38:20] try is something that we call a tokamak.
[38:23] So, the tokamak takes a lot of the
[38:24] concepts, you know, twisting magnetic
[38:26] fields, certain ways of stabilizing some
[38:28] of these instabilities using external
[38:30] coils
[38:32] to create and confine a plasma.
[38:35] And when the first tokamak was basically
[38:38] sort of created, it immediately
[38:40] outperformed almost any fusion concept
[38:42] to date.
[38:43] That was quick done in the USSR, that
[38:45] was quickly verified by Western
[38:47] scientists, and the tokamak since that
[38:49] point has become the dominant focus of
[38:51] fusion research. And it's pretty clear
[38:53] that you can see why. You know, we look
[38:55] at our ion temperature versus Lawson
[38:57] plot, you know, and tokamaks pepper the
[38:59] middle of this space, and in fact, come
[39:01] very, very close to breaking the Q
[39:04] equals 1 boundary, okay?
[39:06] We have an enormous amount of experience
[39:07] building these devices. You know, 170
[39:10] devices have been built. Six of those
[39:12] have been built at MIT over the years.
[39:14] Extensive physics, very good
[39:16] technologies, and they're the only
[39:17] devices
[39:18] that have basically got there, okay?
[39:21] So, if we were, you know, this was
[39:23] probably a lot of concepts. We went
[39:24] through a lot. It's a lot of
[39:25] information, but it's important to see
[39:27] the myriad of ways that people have
[39:29] tried to create the conditions for
[39:31] fusion.
[39:32] So, let's summarize. So, let's look at
[39:34] non-tokamak approaches, okay? So, in our
[39:36] space, every single non-tokamak approach
[39:39] occupies this area. You know, and
[39:41] unfortunately, if your hope is to take
[39:43] these devices and in 2 years make fusion
[39:45] energy, you know, you're not quite there
[39:47] yet, right? You're not tall enough to
[39:49] ride. Maybe you'll get there, you know,
[39:51] maybe you never will because there are
[39:52] instabilities, but the reality is you
[39:55] are far from the place that you need to
[39:56] be to make the conditions for fusion.
[40:00] The tokamak, however, has gotten there.
[40:03] Now, ITER, which some of you may have
[40:05] heard of, ITER is the world's largest
[40:07] science experiment now under
[40:08] construction.
[40:09] Uh it's being built by basically every
[40:12] technologically capable nation on Earth.
[40:15] Uh you can see this device here. It's a
[40:16] tokamak. There's a plasma in the middle.
[40:18] You know, these few pixels are actually
[40:19] an ITER person, so you can see the size
[40:21] of this.
[40:22] One way to look at ITER is to say that
[40:25] ITER is a $50 billion statement of
[40:27] conviction about the tokamak by the
[40:31] world's leading scientific nations and
[40:32] institutions that the tokamak and the
[40:34] tokamak alone has achieved sufficient
[40:36] performance and it's ready to tackle
[40:38] producing electricity from net fusion
[40:41] energy.
[40:42] Okay? And this is backed up, you know,
[40:44] national academies, blue ribbon reports,
[40:46] you know, et cetera.
[40:48] It's an important point, however, to say
[40:51] that the tokamak's physics achievements
[40:54] do not validate, nor do they elevate
[40:56] non-tokamak claims to be nearing the
[40:59] production of fusion energy. Okay? So,
[41:01] just because the tokamak overcame many,
[41:04] many instabilities, you know, through 60
[41:06] years of research and billions of
[41:07] dollars of funding, they might have
[41:09] solved all their own problems, but they
[41:11] didn't solve a single other problem for
[41:12] these other devices. They're stuck still
[41:15] having to figure out a way, if it's
[41:16] possible, to climb up towards these
[41:18] conditions. It's not to say they can't
[41:20] do it, but it is to say that the tokamak
[41:22] doesn't help them. Okay? They continue
[41:24] to be very interesting scientifically,
[41:26] but they have a lot of progress to go to
[41:27] sort of demonstrate energy relevance on
[41:29] the near term.
[41:30] So,
[41:31] coming back to where we started uh in
[41:33] this analogy with with flight, right?
[41:35] Remember the the the picture of the
[41:37] plane, right? Yes, you know, where's
[41:38] your wing? That's your key question. So,
[41:40] your key question in fusion, when you
[41:42] evaluate a fusion concept, somebody
[41:44] comes to you and they says, "I have this
[41:45] amazing reactor design." Right? I
[41:47] particularly like Mr. Fusion from Back
[41:49] to the Future. Um you know, but your key
[41:51] question should be, "What's your
[41:52] temperature? What's your density and
[41:54] your energy confinement?"
[41:55] That is essentially what you need to
[41:57] know to know how close a particular
[41:59] concept is to demonstrating the ability
[42:01] to make electricity from fusion. So, we
[42:03] can finally, after all that, answer
[42:05] question three.
[42:07] What fusion energy approaches exist? How
[42:09] should they be evaluated? Many
[42:11] approaches to fusion energy have been
[42:12] and are being tried. Only the tokamak
[42:15] has demonstrated the right level of
[42:17] performance. So, the tokamak,
[42:18] you know, with a maximum performance of
[42:20] about 0.65, you know, leads by a lot.
[42:23] It's not there yet. There might still be
[42:25] problems that get encountered, but it's
[42:28] the closest by a lot. Stellarators come
[42:30] next. You know, they'll be a very
[42:31] interesting concept to follow. All the
[42:33] others, you know, are a factor of 10,000
[42:35] to a million times behind the tokamak in
[42:38] the key physics metric.
[42:41] So, you have questions, you know, that
[42:42] you can ask. You know, has this
[42:44] particular approach to fusion been tried
[42:45] before? Why was it previously abandoned?
[42:48] Why did you pick it up? What changed?
[42:49] How far
[42:51] in, you know, density, temperature,
[42:53] confinement time are you extrapolating
[42:55] from where you are today to show energy
[42:57] relevance? How do you propose achieving,
[43:00] you know, this factor of 10,000 to a
[43:01] million times performance increase that
[43:04] you need to be energy relevant? How long
[43:05] is that going to take? At what cost? You
[43:07] know, that's really important question
[43:10] that need to be asked of different ways
[43:11] of making fusion energy, including of
[43:13] the tokamak.
[43:14] So, here's our rules just to recap.
[43:17] Rule one is fuel choice fundame-
[43:19] fundamentally sets the difficulty. Okay?
[43:21] Rule two, the closer you are to a
[43:24] burning plasma, that is the ultimate
[43:26] arbiter of the viability of your
[43:27] approach to fusion energy. You know,
[43:29] rule three is that many approaches to
[43:31] fusion energy have been and are being
[43:32] tried. Only the tokamak has demonstrated
[43:35] the right level of performance.
[43:38] So, now let's take a look given what we
[43:40] know about MIT's proposed pathway, what
[43:43] are we proposing to do that's different
[43:45] to get to fusion energy quick?
[43:47] So, we said that the tokamak is this
[43:49] great concept. It's outperformed
[43:50] everybody. Okay? But we should point out
[43:53] that not all is well with tokamaks.
[43:55] Okay? Tokamaks have serious problems.
[43:57] So, what's the problem? So, let's plot
[43:59] the density, the temperature, and the
[44:01] confinement time on the Y axis, and you
[44:04] know, time on the X axis. So, you can
[44:05] see here tokamaks quite quickly marched
[44:09] all the way up to where they need to be.
[44:11] Till about mid-90s. And then progress
[44:14] stopped, ceased entirely in this
[44:16] particular parameter. And you should ask
[44:18] those questions, you know, what caused
[44:20] your performance to stop? Well, did we
[44:22] hit some performance limit? You know, we
[44:24] we simply can't go any faster or any
[44:26] harder?
[44:27] Uh did we encounter some insurmountable
[44:29] instability we couldn't figure out how
[44:31] to fix? Uh was there something just
[44:33] totally unknown in physics that we
[44:34] didn't, you know, think was there that
[44:36] scuppered our our ability to increase
[44:38] this key parameter?
[44:40] The answer is no.
[44:43] Physics performance ceased in the
[44:45] tokamak due to size. It did not saturate
[44:48] due to any physics problem.
[44:50] So, JET, which is the device that made,
[44:53] you know, the the the peak performance
[44:54] of a tokamak, here you can see it in
[44:56] 1997 was when it did that. You know, the
[44:58] next proposed device is ITER. You know,
[45:01] again, these few pixels are a human.
[45:04] Its peak performance is expected maybe
[45:06] around 2040. Um and you can sort of see
[45:09] that this is a big step, right? If we
[45:11] were to look up on our curve and put the
[45:13] point of ITER on here, you can see why
[45:15] there's this very strong dog leg. It's
[45:18] simply size. And if you take the ITER
[45:20] approach to fusion energy, this is what
[45:22] it looks like. Okay? You start here
[45:24] today with devices that we have. We'll
[45:26] open a new device in a few years. We
[45:29] will build ITER sometime around 2030, 20
[45:32] you know, we actually start to run ITER
[45:33] at its peak performance around 2035,
[45:35] 2040. There are other devices
[45:38] envisioned, you know, fusion nuclear
[45:39] science facility, and then the Europeans
[45:41] have a plan to build a demonstration
[45:43] reactor sometime after 2050. Okay?
[45:46] So,
[45:47] this graphic to me embodies the typical
[45:51] tokamak critique, which many people are
[45:52] familiar with. Tokamaks are too big,
[45:55] they are too complex, and they are too
[45:56] slow to be relevant to any issue in the
[45:59] near term on climate change and energy
[46:01] production issues.
[46:03] I completely agree with this statement.
[46:05] But there is a subtlety that we need to
[46:08] recognize. So, the reason that this
[46:10] pathway looks like this was caused by
[46:12] people, people in technology making
[46:14] decisions on what tokamaks, what types
[46:16] of tokamaks we would build.
[46:18] It was also caused by the organizational
[46:20] complexity that's required to organize
[46:23] things that are this size. And that this
[46:26] is not a reason to abandon the tokamak.
[46:28] So, just because ITER and this pathway
[46:31] is large, it's expensive, it's long
[46:32] time, doesn't mean it won't work.
[46:34] This is not a reason to abandon the
[46:36] tokamak.
[46:38] So,
[46:39] turns out that magnetic fields are very
[46:42] important in demonstrating uh in sort of
[46:44] affecting the size of a particular
[46:46] tokamak. So, why is that? Well, you
[46:48] know, most simply come back to this idea
[46:50] of confinement. You know, we use
[46:51] particles uh magnetic fields to confine
[46:53] particles.
[46:55] Uh and you know, the confinement
[46:56] improves as you increase the size of B.
[46:59] You know, your particles, as you
[47:00] increase the B field, gyrate tighter and
[47:02] tighter around these lines. Okay? So,
[47:04] your confinement gets better. Your and
[47:06] your plasma, your energy confinement
[47:08] becomes better insulated.
[47:10] You know, the plasma becomes more
[47:11] stable, right? Uh and the magnetic
[47:14] pressure, you know, again, this density
[47:16] times temperature factor
[47:18] uh goes like the magnetic field squared.
[47:20] So, if you think about a tokamak, the
[47:21] plasma is expanding, magnetic fields are
[47:23] pushing on it, and magnetic pressure,
[47:25] which was one of those key parameters we
[47:26] looked at before, uh is highly dependent
[47:29] on the magnetic field.
[47:30] And then the fusion reactivity, you
[47:32] know, how many fusions you're making,
[47:34] you know, per unit volume,
[47:35] is very, very strongly correlated with
[47:37] the magnetic field. It goes like the
[47:39] magnetic field to the fourth. So, things
[47:41] like energy gain, you know, this
[47:43] parameter, goes like magnetic field to
[47:44] the third. The power density, you know,
[47:47] how much power are you making per cubic
[47:49] meter of stuff, let's say, which is a
[47:51] critical quantity for economic
[47:53] viability, goes like the magnetic field
[47:56] to the fourth. It's a really strong
[47:58] dependence that we shouldn't ignore.
[48:00] So, let's take a look at what this
[48:02] implication of magnetic field has for
[48:03] the tokamak.
[48:05] If we were to plot the magnetic field on
[48:07] the X axis,
[48:08] plot the size of the device on the Y
[48:10] axis, because we've understood the
[48:12] physics of tokamaks, we can actually
[48:14] plot these contours in this space. So,
[48:17] these contours are contours of constant
[48:20] gain or capital Q. So, you know, ideally
[48:23] you want to march right up these
[48:24] contours into a place where you're
[48:25] generating net energy.
[48:27] An important thing to understand is that
[48:29] if you were in this region of size,
[48:32] you're at the global scale. You are
[48:33] ITER. You require many nations to
[48:35] contribute to building a device. If
[48:37] you're sort of here in this range,
[48:38] you're kind of at the national lab
[48:40] scale.
[48:41] Uh and if you're down here, you're sort
[48:42] of at the university or the, you know,
[48:44] small company scale. Those are the
[48:46] things that you can afford to do uh at
[48:48] that size.
[48:49] So, what was the problem with ITER? You
[48:52] know,
[48:53] it's not that ITER
[48:55] uh you know, made bad decisions. In some
[48:57] sense, they were constrained by the
[48:59] technology available. Okay? So, with the
[49:01] magnetic fields that could be produced
[49:04] uh with something called a
[49:05] superconductor, right? Superconductor is
[49:07] a material that conducts uh perfectly,
[49:10] does not lose any energy to resistivity.
[49:12] That's nice because you can make a
[49:13] magnet that's always on, doesn't require
[49:15] power other than to keep it cold. So,
[49:16] that's nice.
[49:18] So, the superconductors that existed
[49:19] when the ITER design was fixed,
[49:22] basically let you go to about, let's
[49:24] say, 5 and 1/2 Tesla. You simply
[49:26] couldn't push the magnetic fields any
[49:28] higher. There's actually a fabulous uh
[49:30] statement in the ITER design document,
[49:32] which to paraphrase says, "We understand
[49:34] the physics of tokamaks. We know the
[49:36] importance of magnetic field. We will go
[49:38] to the highest magnetic field the
[49:39] technology lets us go to, and then we
[49:42] will simply have to fix our size to get
[49:44] the performance up.
[49:45] There's nothing they could do. So, if
[49:47] you want to build a reactor with
[49:48] traditional superconductors, you are
[49:50] constrained to build in this very small
[49:52] wedge because you cannot push to higher
[49:54] magnetic field. So, you might be able to
[49:55] get a net energy gain of 20, but your
[49:58] size is enormous.
[50:00] And so, one of the questions we've asked
[50:02] at the Plasma Science and Fusion Center
[50:03] is has anything changed? Has anything
[50:06] changed in the last 10 years or even 5
[50:09] years that might let us get rid
[50:11] of this constraint on magnetic field?
[50:13] And the answer is yes.
[50:15] The answer is this thing called
[50:16] high-temperature superconductors, okay?
[50:19] So, what are these things? So, these are
[50:21] superconductors that let you expand the
[50:24] parameter space of when they are
[50:26] superconducting. So, a superconductor is
[50:28] defined by temperature that it operates
[50:30] at, the magnetic field it sits in, and
[50:32] how much current you're trying to push
[50:34] through it. The superconductors, low
[50:36] temperature that ITER designed, you can
[50:38] see in these red curves. You know, it's
[50:39] very very small operating space. They're
[50:41] really cold. They can't make much
[50:43] magnetic fields, otherwise they go
[50:45] normal conducting. That's bad for
[50:46] superconducting magnets. HTS expands the
[50:49] space. Lets you do things you simply
[50:51] couldn't think about doing before. So,
[50:53] what does that mean? Well, if you can
[50:54] make much higher magnetic field magnets,
[50:56] you can either reduce the fusion size
[50:58] the the size of the device for the same
[51:00] performance, or at the same size you
[51:02] could drastically increase the
[51:03] performance.
[51:05] You can operate these things at higher
[51:06] temperatures. So, now you have new
[51:07] cryogenic options for keeping the
[51:09] superconductor cold. You have better
[51:10] material properties. You have higher
[51:13] current densities, so you can think
[51:14] about making magnets that have that are
[51:16] smaller, they're more compact, and they
[51:18] have more structure, so they can take
[51:19] more load.
[51:21] An important thing to point out is that
[51:23] HTS, high-temperature superconductor, is
[51:25] not a benchtop curiosity at this point.
[51:28] So, in the last few years, HTS has moved
[51:31] from being a benchtop curiosity to an
[51:32] industrially produced product. You can
[51:34] go out and buy it, and it has sufficient
[51:36] performance today off the production
[51:38] lines that you can make magnets for a
[51:40] high-field fusion.
[51:42] And there's a whole number of companies
[51:43] involved in making all of this tape.
[51:46] It's really good stuff. So, we've done a
[51:48] lot of work with this at the PSFC. You
[51:50] know, and HTS magnets are a reality, and
[51:52] these are a real thing.
[51:54] So, what does that let us do? So, let's
[51:55] go back to our magnetic field versus
[51:57] size space. We've now removed this
[51:59] constraint. This lets us think about
[52:01] building reactors at much higher field,
[52:04] but at much smaller size.
[52:06] So, a few years ago, there was a design
[52:09] study that's more or less ongoing. It
[52:11] was done by a class under Dennis White
[52:14] and the Nuclear Science and Fusion
[52:15] Center. So, this was a device
[52:17] designed by graduate students. It was
[52:19] called ARC. And in the space, you know,
[52:21] ARC took these HTS magnets and said,
[52:23] "What can we do with them?"
[52:25] Boom. So, they put a point down in this
[52:27] space. So, ARC is about 3 m in major
[52:30] radius. That's the distance from the
[52:31] center
[52:32] to the plasma. You know, and operated at
[52:35] high magnetic field, something around
[52:36] 9.2 T.
[52:38] Well, let's compare that to ITER. Does
[52:40] it make things better? So,
[52:42] if I want to make 500 MW of fusion
[52:45] power, which is the same thing ITER
[52:46] does, I can do that in ARC with the
[52:49] exact same physics that the world has
[52:51] basically signed off on will work. And I
[52:54] could do that at a device which is about
[52:56] 10 times smaller.
[52:58] Cheaper to build,
[52:59] shorter time scales,
[53:01] less institutional complexity to
[53:03] organize.
[53:04] Now,
[53:06] this argument, you know, doesn't stop
[53:07] there. So, if higher magnetic fields
[53:10] enable ARC to rethink how fusion energy
[53:12] tokamaks, you know, are designed, why
[53:13] would you stop?
[53:15] Just keep going. So, is there an
[53:16] interesting space over in this region
[53:19] where you have very high magnetic
[53:20] fields, and you are now most importantly
[53:23] down at a university scale device?
[53:26] And so, there's an ongoing, you know,
[53:28] serious design effort at the PSFC to
[53:30] think about what a device would look
[53:31] like that could do this. We call it
[53:33] SPARC. SPARC stands for smallest
[53:35] possible ARC.
[53:37] And what this lets us think about doing
[53:38] is building the world's first net energy
[53:40] device with incredibly robust and mature
[53:44] physics with a key piece of innovation,
[53:46] which is the high high-field magnets
[53:48] enabled by HTS. And you can do that, you
[53:51] know, down here. University scale. You
[53:54] don't have to organize the world's most
[53:56] advanced nations to build a $50 billion
[53:58] experiment. That's a really compelling
[54:00] argument for why you build high-magnetic
[54:02] field tokamaks. It's the confidence that
[54:04] we have in the physics.
[54:07] Okay? So, ultimately, you know, we think
[54:09] if climate change is a problem, if
[54:11] energy production issues are a problem,
[54:13] if where you get that energy is a
[54:15] problem,
[54:16] we need more options, right? And if
[54:18] those things are important in the near
[54:20] term, you know, this pathway might work
[54:22] using tokamaks.
[54:23] But we think by applying a key piece of
[54:25] innovation, there's a much more
[54:26] attractive pathway. And it goes from a
[54:28] device that you can go see today at 4,
[54:30] which is about this size, you know,
[54:31] person is about here.
[54:33] You go through a device called SPARC,
[54:35] you know, and sometime in the near term,
[54:37] you know, 10 15 years, you could
[54:39] feasibly think about building devices
[54:42] that are actually putting watts on the
[54:43] grid you know, in time to make a
[54:45] difference. Okay?
[54:48] So, I'll just leave a conclusion up
[54:49] here, which is, you know, we tried to
[54:51] take this sort of panoply of fusion
[54:54] concepts, which could be very confusing.
[54:56] What's the right question to ask? Well,
[54:58] you know your question. You know, what
[54:59] does your wing look like? In this case,
[55:01] it's what is your
[55:03] loss in parameter and what's your
[55:04] temperature. You know, and keep this
[55:05] picture in your mind of when you plot
[55:07] those things in the space, right? There
[55:09] are concepts, the tokamak, which are
[55:11] clearly orders of magnitude in advance
[55:14] of some of these other concepts. And if
[55:15] doing things on a short time scale with
[55:17] high confidence is what we need to do,
[55:20] tokamaks look pretty good. And the
[55:22] pathway that's being envisioned at the
[55:23] PSFC takes us from where we are today
[55:26] and goes on a short time scale to
[55:27] something that puts watts on the grid.
[55:29] THANKS.
[55:43] WE HAVE TIME FOR QUESTIONS. I THINK YOU
[55:45] CAN JUST
[55:46] SURE. And so, if anybody has any
[55:48] questions or any comments, In the back.
[55:50] I'd like to ask a question about your
[55:52] fourth
[55:53] item there on your um
[55:55] fusion reactions or on
[55:58] proton capture.
[56:01] A few years ago here at MIT, they had a
[56:03] project to treat brain tumors
[56:06] using a boron neutron capture
[56:09] method. And I'm just wondering what the
[56:12] difference is. I mean, apparently they
[56:14] solved the problem of getting
[56:16] boron into brain tumors, and somehow
[56:19] they managed to convince patients to
[56:21] slide under the nuclear reactor. It's
[56:24] interesting, but what exactly was the
[56:25] process? Was that a fusion or a fission
[56:29] reaction? It's So, that is an important
[56:32] detail. Let me find my table.
[56:35] So, something that to, you know, there's
[56:37] nothing
[56:38] in principle special about a fusion
[56:39] reaction. So, any rearrangement of the
[56:41] nuclei we would we would sort of more
[56:43] generally just call a nuclear reaction.
[56:45] So, whether it's, you know, two
[56:46] deuteriums coming together, fusing, and
[56:48] other stuff comes out, whether it's a
[56:49] neutron coming in, you know, and that's
[56:52] sort of a nuclear reaction. So, there is
[56:53] a process by which neutrons can be
[56:56] absorbed into
[56:58] nuclei. And typically, those reactions
[57:01] have very high energy gain. So, you get
[57:03] high energy output from those reactions.
[57:06] And more importantly, the cross-section,
[57:09] the probability that neutrons will, you
[57:12] know, be absorbed into certain nuclei
[57:14] are some of the highest probabilities
[57:16] for any known nuclear reaction in the
[57:18] universe.
[57:19] So, that's actually a really good
[57:20] question. Why don't you use neutrons?
[57:21] So, when I actually when I took this
[57:23] from Todd Rider, he actually has a cat
[57:25] another column here, which is neutrons.
[57:28] And it turns out there are actually
[57:29] quite a few number of reactions that if
[57:31] you just look at it, you say, "Oh, this
[57:33] looks great. You know, high
[57:34] cross-section, you know, good
[57:35] energetics." Clearly it happens. You
[57:37] know, we use those reactions all the
[57:39] time.
[57:40] The rub comes in into producing
[57:42] neutrons. So, it turns out that actually
[57:44] making neutrons is an enormously
[57:47] energy-intensive affair.
[57:49] And so, if your idea is to make
[57:51] neutrons,
[57:52] then use reactions to create
[57:54] electricity,
[57:56] you still have to feed that electricity
[57:58] back in to make neutrons. So, that's
[58:00] actually something we call recirculating
[58:02] power, which is a really important
[58:04] metric for how efficient is a fusion
[58:07] device.
[58:08] And it turns out if you have to make
[58:09] neutrons, you're never going to be able
[58:12] to sort of make, you know, your
[58:13] recirculating power anywhere near, you
[58:15] know, feasible for something that puts
[58:17] net energy on the grid.
[58:25] Other questions? In the back.
[58:28] I
[58:28] have a question about the
[58:29] high-temperature superconductor.
[58:31] HTS was discovered in the '80s.
[58:34] The early '90s was a loss of excitement
[58:36] for them. Yep. It was a loss of the
[58:38] world actually doing
[58:40] high-temperature superconductors. So,
[58:41] aluminum replaced niobium, and so forth.
[58:44] And then the concept was sort of
[58:45] abandoned, and everyone went back to the
[58:47] niobium interest for conductivity sake.
[58:50] So, what is
[58:51] So, why was it abandoned back then? And
[58:54] why was it abandoned?
[58:56] So, when when what
[58:57] we call them rare earth barium copper
[58:59] oxides is the most, you know, leading
[59:01] candidate. The highest performing ones.
[59:03] When they were discovered, they were
[59:04] discovered as these sort of little
[59:05] interesting grains, you know, that sat
[59:07] on somebody's benchtop. You know, and
[59:09] you could sort of make them, you know,
[59:10] do superconductivity. They could go
[59:12] superconducting at incredibly high
[59:14] temperatures. So, there there's actually
[59:16] a moment in physics that was called the
[59:17] Woodstock of physics. It was not nearly
[59:20] as fun and debaucherous as the actual
[59:21] Woodstock, but you know, when they
[59:23] demonstrated this at at a big major
[59:25] conference, people were up on a table,
[59:26] and they were looking at this stuff, and
[59:28] it basically you sort of rocked the
[59:30] world of superconductivity. So, the
[59:32] problem was taking those tiny little
[59:34] crystals, which were interesting, and
[59:36] turning them into something that you
[59:37] could industrially produce at a large
[59:39] scale. In a format, you know, in a form
[59:42] factor that you could use to do real
[59:43] engineering if you wanted to build
[59:44] magnets.
[59:45] So, it it took about 20 years for
[59:48] industrial, mostly deposition processes
[59:51] to occur that allowed you to make these,
[59:54] you know, REBCO materials, which have a
[59:56] very definitive grain structure.
[59:59] Um you had to develop the technology
[60:00] that not only could
[60:02] do certain types of material uh
[60:04] deposition, you know, with the right,
[60:06] you know, uh atomic structure, but you
[60:08] also had to be able to do that at scale.
[60:10] And so, only in sort of the last 5 to 10
[60:12] years have these processes, they're
[60:13] called IBADs or RABiTS, you know,
[60:15] different words,
[60:16] um different types of processes, but
[60:18] basically only in the last 5 years have
[60:19] those gotten to the point where you can
[60:21] now set up these large, sort of
[60:23] large-scale manufacturing lines. And
[60:25] now, you can, you know, make these, you
[60:27] know, they're tapes, basically. So,
[60:28] they're sort of these long, think of
[60:30] them as like long steel ribbons. Now,
[60:32] you can actually make this stuff in
[60:33] kilometer quantities, and it's useful,
[60:34] and you can do stuff with it. And the
[60:36] performance has only continued to
[60:37] increase.
[60:38] So, it's really a manufacturing
[60:40] innovation that took it from a benchtop
[60:41] curiosity to something that's an
[60:43] industrial-scale product.
[60:51] So, assuming that, you know, the tokamak
[60:53] works out and you can achieve those Q
[60:55] values, are there any major engineering
[60:57] problems coming in getting that power
[60:58] out of a tokamak and onto the grid? Do
[61:00] you see any big hurdles there that would
[61:02] increase that timeline? Well, so, in in
[61:05] principle, you know, taking neutrons
[61:09] and making electricity
[61:11] uh was discovered in the, you know, '40s
[61:13] and '50s.
[61:14] So, you know, we've known for a long
[61:16] time, you know, uh there's a great
[61:18] picture, actually, if you want to go and
[61:19] and look at it. This was out in Idaho
[61:21] National Labs. It's the first time that
[61:23] anybody had ever put electricity on the
[61:24] grid from fission. I think it was at
[61:26] EBR-1 is the reactor, and they strung up
[61:29] four light bulbs,
[61:30] uh and they actually glowed. So, the
[61:32] first time in maybe a millennium that
[61:34] humanity had invented a new energy
[61:36] source from fission and actually made
[61:38] electricity from it. So, in some sense,
[61:40] you know, that technology has existed
[61:42] for a long time. You know, use neutrons,
[61:44] heat stuff, drive steam, turn a turbine.
[61:47] Now, there are subtleties, right? So,
[61:48] neutrons damage materials. So, if you
[61:50] have a high fluence of neutrons coming
[61:52] in, they do damage materials. It's
[61:54] radiation damage, so those are
[61:55] questions. Um you know, what kind of
[61:57] material science do you need? What kind
[61:59] of plasma science do you need? Uh how do
[62:01] you put those things together, you know,
[62:03] in a device of this complexity to make
[62:05] it all work out? Um for a
[62:07] deuterium-tritium fuel cycle, you have
[62:09] to also take those neutrons and and make
[62:11] more tritium. So, you have to also breed
[62:13] fuel.
[62:14] Tritium is not something that is widely
[62:16] available. It doesn't exist naturally on
[62:18] Earth in large quantities, and so you
[62:19] actually have to make it. So, you have
[62:21] to ensure that for each neutron that
[62:23] comes into your material, you're making
[62:25] at least one or more tritium fuel from
[62:28] that reaction.
[62:29] Uh and those are sort of active areas,
[62:31] those are two active areas uh sort of
[62:33] research now, you know, net tritium
[62:35] breeding um and how you deal with some
[62:37] of the radiation damage issues that
[62:39] comes from using neutrons.
[62:43] So, is that getting back to basic basic
[62:45] physics?
[62:47] One thing you didn't mention at all were
[62:48] the
[62:49] uh issue of steady-state operation.
[62:51] Yeah. And especially when you're
[62:52] comparing tokamaks to stellarators.
[62:54] Yeah.
[62:55] I want to know you
[62:57] principle physics principle
[62:59] can be steady-state
[63:02] you know,
[63:03] the simplest quarter
[63:04] not steady-state. Mhm.
[63:06] And you really like steady-state. So,
[63:08] you're saying this
[63:09] And that that particularly impacts
[63:10] tokamak versus stellarator.
[63:12] Yep. So, do you want to say elaborate on
[63:14] that?
[63:15] You you know, to me,
[63:17] uh you know, I I always come back to
[63:19] this idea of, you know, how close are
[63:21] we?
[63:21] So, you know, the stellarator again has
[63:23] advantages, tokamak has advantages, and
[63:25] they both have disadvantages. You know,
[63:26] to me, you know, I'm curious to know
[63:29] if we can actually make net energy in
[63:31] the near term.
[63:32] Because I think the instant that you
[63:34] have a device
[63:36] which makes net fusion energy,
[63:39] everything changes. It doesn't actually
[63:40] matter, in some sense, what some of the
[63:42] engineering difficulties, maybe the
[63:43] plasma difficulties are at that point.
[63:46] So, you've now taken something which was
[63:47] a problem,
[63:48] uh i.e. my tokamak isn't steady-state,
[63:52] and you've now made it an opportunity.
[63:53] If I can make my tokamak run
[63:54] steady-state,
[63:55] wow. I mean, now I actually, you know,
[63:57] I'm ready. I I can engineer, I can do
[63:59] all these things. I think the idea of
[64:01] running a tokamak steady-state is a
[64:02] really critical uh issue for
[64:05] uh actually this pathway that we're
[64:07] developing the PSFC. Um and it's, you
[64:09] know, you know, we have a lot of
[64:10] materials, and that is actually
[64:12] uh a science problem, a physics problem
[64:15] that we are directly intending to
[64:16] tackle. So, you know, as you know, you
[64:18] know, you could imagine converting the
[64:19] device that we have today, Alcator, uh
[64:22] into a a small, fast, you know, cheap
[64:24] experiment that would actually prove
[64:27] that you can demonstrate, you know, the
[64:29] physics of driving, you know, current in
[64:31] a plasma making it steady-state in
[64:32] tokamak. And we should point out, you
[64:34] know, that at Alcator C-Mod, where we
[64:36] have a robust physics program in just
[64:38] this area, we have demonstrated that you
[64:40] can drive current, you know, in a plasma
[64:43] in steady-state, theoretically
[64:45] infinitely, using RF waves.
[64:47] So, some of the physics still has to be
[64:49] worked out, but at least the basic
[64:50] principles, you know, are being put in
[64:52] place today, you know, on machines, and
[64:54] we've demonstrated the ability to make
[64:57] tokamaks run in steady-state. At least,
[64:59] you know, it's a glider at this point,
[65:01] right? Does it scale? We don't know, but
[65:03] the I think the initial the initial
[65:05] inclination is good.
[65:11] Do you already have a
[65:12] budget?
[65:15] Are you Do you have a big pocketbook?
[65:20] Uh for ARC, no. So, you one of the ways
[65:22] to think about ARC, which I think is a
[65:24] really good one, um is an interesting
[65:26] one being in a university, is what
[65:27] happens when you turn students loose
[65:30] uh without preconceptions about what
[65:31] can't be done. So, that's actually one
[65:33] of the interesting outcomes of ARC um
[65:35] was that students sort of took this
[65:37] project and made it an interesting um
[65:39] design study.
[65:41] But, you know, you shouldn't think of
[65:43] ARC as something which is ready to build
[65:44] tomorrow, right? So, I'm not ready to
[65:46] take money from you and to start pouring
[65:48] concrete to make ARC. You know, what
[65:50] what you should think about ARC is uh
[65:52] ARC is sort of a
[65:53] conceptual design for what happens in
[65:56] tokamaks when you apply this single
[65:58] innovation. So, if I take HTS
[66:01] and I think about the tokamak
[66:03] and I think about making it different
[66:05] from ITER. Now, what does it let me do?
[66:07] And it turned out it came with, you
[66:07] know, ARC came up with all these
[66:09] different innovations, which solved a
[66:10] lot of problems that I didn't have time
[66:11] to go into. So, you know, ARC is sort of
[66:14] a a model of what we think a tokamak
[66:16] reactor might look like. There's a lot
[66:17] of compelling um
[66:19] sort of design principles embedded in it
[66:21] and and ways to approach things, but you
[66:23] know, there's no guarantee that when we
[66:25] build SPARC, we get fusion energy gain,
[66:27] great, um and then we're ready to build
[66:29] something that puts watts on the grid.
[66:31] It may not look exactly like ARC. But,
[66:33] it was certainly a very interesting
[66:34] exercise to sort of flesh out some of
[66:36] the basic physics of high magnetic
[66:38] fields, uh to do a lot of the sort of
[66:40] rough engineering for what it might look
[66:42] like to take a a DT tokamak with very
[66:45] high magnetic fields uh with this new uh
[66:48] material and actually, you know,
[66:51] make a device that has net energy gain.
[66:53] But, obviously, someday we want to build
[66:55] something that looks like ARC in the
[66:57] near term. Whether it looks exactly like
[66:59] that, we don't know.
[67:00] We're not actually soliciting
[67:02] contributions for ARC at the moment.
[67:08] So, you made the point that by going to
[67:09] HTS, you can double the magnetic field
[67:11] and that
[67:13] greatly shrinks the size of the tokamak.
[67:16] But, the technology is not static, it's
[67:17] getting better and better. So, my
[67:18] question is,
[67:20] what happens if they're able to double
[67:22] again the magnetic field in HTS? Would
[67:24] it would the tokamak get even smaller or
[67:26] do you do you
[67:28] encounter other other limits? Right. So,
[67:30] that, you know, that would be nice.
[67:31] Let's just keep boosting the magnetic
[67:33] field up. So, if you look at this space,
[67:34] for instance, um you know, there are
[67:36] some
[67:38] You know, from the physics standpoint,
[67:39] physics just seems to keep getting
[67:41] better with magnetic field. Looks great.
[67:43] You know, the problem is, you now run
[67:44] into engineering problems. So, your
[67:47] magnetic field keeps going up, uh and
[67:49] now you run into, let's say, two
[67:51] problems. So, one problem
[67:53] um that we talked about is that you have
[67:55] to take the neutrons in a deuterium
[67:57] fusion reactor, and you have to do
[67:58] something useful with them.
[68:00] And so, you know, neutrons at a specific
[68:03] energy, in the in this case, the ones
[68:04] that come out of DT go a certain length
[68:07] uh through materials, okay? And so, you
[68:10] have to make sure uh that you have
[68:12] enough materials that you could, you
[68:14] know, capture those neutrons. So, you
[68:15] might try
[68:17] uh to take a reactor
[68:19] uh and make it, you know, 14 Tesla, you
[68:21] know, and you're down here, you're sort
[68:22] of
[68:23] at half a, you know, a meter scale,
[68:25] okay? That might look good. There might
[68:27] be other physics issues there, but now
[68:29] you sort of haven't left enough room if
[68:31] you've tried to shrink your device with
[68:32] magnetic fields to do something useful
[68:34] with the neutrons. You might be able to
[68:35] get away from that with some other type
[68:37] of fuel cycle, but that could be
[68:38] interesting. Um
[68:40] but sort of a more pressing issue is, do
[68:42] your magnets break?
[68:44] So, you end up, you know, as you put
[68:46] more current into these magnets, uh and
[68:48] they generate, you know, uh currents in
[68:51] the in the coils, those coils are, you
[68:53] know, next to other magnets, they're
[68:54] under immense structural loads. And so,
[68:57] one of the key limits uh in a fusion
[69:00] device for how big you can set your
[69:01] magnetic field is, can you simply
[69:03] engineer magnets that can tolerate the
[69:05] structural loads of going to higher
[69:07] field? And that almost sort of sets what
[69:09] we think is the practical limit at
[69:11] present for how high a magnetic field
[69:13] you could go. So, do your magnets just
[69:16] Are they able to take the structural
[69:17] load?
[69:23] Chances.
[69:28] energy
[69:29] uh very interesting topic and everybody
[69:33] every moment everywhere we use the
[69:35] energy to fuel our car and
[69:38] etc.
[69:39] Um and um
[69:41] the tokamak low pollution not like
[69:43] burning a coal like low pollution.
[69:46] So um tokamak sounds a very promising
[69:50] path to solve problem of energy.
[69:53] Um
[69:55] In your like if you just pick up one key
[69:58] issue to solve problem
[70:00] only one
[70:01] uh
[70:02] your opinion what is it? Is it like um
[70:05] scientifically or engineering?
[70:08] So it is like scientifically it's a well
[70:11] understood and then we need uh people to
[70:14] work Right. So
[70:16] your problem. For example, the material.
[70:18] Now we use the high temperature
[70:20] superconducting materials or other
[70:22] techniques. So what is the one?
[70:26] So
[70:27] uh
[70:27] I have So
[70:29] I think the number one problem that
[70:30] needs to be solved in fusion is somebody
[70:32] has to put a point here.
[70:35] That's it.
[70:37] I honestly believe that. The first
[70:38] person the first group the first
[70:40] institution the first government whoever
[70:42] it is the first person who puts a point
[70:45] in this space
[70:46] changes everything. And that's going to
[70:48] involve plasma physicists it's going to
[70:50] involve structural engineers mechanical
[70:52] engineers it's going to involve
[70:53] electrical engineers you know all sorts
[70:55] of things. But I actually think that
[70:57] right now in fusion
[70:59] the biggest problem in fusion is not
[71:00] technology
[71:02] or physics. The biggest problem in
[71:04] fusion is perception. It's the
[71:05] perception that fusion's a joke
[71:08] that it's not serious that it's 30 years
[71:10] away and always will be that this is not
[71:12] a real contributor to energy in the near
[71:14] term. That I think is the biggest
[71:16] problem for fusion.
[71:18] And the instant you put a point here
[71:21] it completely changes the story.
[71:31] Okay thank you for the talk.
[71:32] No more questions let's thank our
[71:34] speaker. Great thanks everyone.