Scientific Gain and Fusion Power Plants

“For a successful technology, reality must take precedence over public relations, for nature cannot be fooled.” - Richard Feynman¹

Introduction

This series of articles explains the physics requirements of a useful fusion energy system, with interactive graphics,² minimal jargon, and no math. By reading this series and interacting with the graphics, you'll develop an intuition for how near or far the physics of a fusion experiment is to a useful fusion energy system (i.e., one that makes net electricity). You'll also learn what questions to ask fusion machine developers in order to separate the physics from the hype.

This article,³ the first in the series, focuses on power flows in conceptual fusion power plants, independent of any one approach to fusion energy. The framework applies to magnetic confinement (e.g., tokamaks, stellarators, and mirrors), inertial confinement (e.g., laser or pulsed power systems), and approaches that utilize both magnetic and inertial concepts together (e.g., compression of magnetically confined systems). The power flows will be explained piece-by-piece, culminating in an integrated power plant model. Future articles will focus on the fusion plasma itself, and related topics.

It's important to note that physics performance is not the only requirement for the widespread deployment of fusion energy. Attractive economics⁴ and social acceptance⁵ are as important as the physics, though this series only covers physics.

Fusion as a power amplifier

In a fusion system, heating energy (in various forms depending on the fusion concept) crosses into a vacuum vessel to heat and sustain a fusion plasma. That original heating energy is eventually released (albeit in a different form)⁶ plus energy liberated by fusion reactions. A time-averaged view of these processes results in the steady-state power flows shown in Figure 1.

Q is a generic term for gain: the ratio of fusion power liberated to the heating power needed to sustain the reaction. Because the magnitude of the heating power depends on where you measure it, we can sharpen things up by defining "scientific gain" Q_{\rm sci} as the ratio of fusion power liberated to heating power crossing into the vacuum vessel. This is a useful definition because it cleanly separates the plasma physics inside the vacuum vessel from the systems outside it. At Q_{\rm sci} = 1, fusion power equals the heating power entering the vacuum vessel. In steady state, what goes in must come out, so the heating power exits alongside the fusion power,⁷ and twice as much total power leaves the vacuum vessel as enters it. Drag the slider to vary Q_{\rm sci} and see the effect.

At Q_{\rm sci} = 10, eleven times as much power exits as enters. Speaking informally, sometimes people say, "Q is the ratio of power out to power in." That's roughly true at large values of gain, but the approximation breaks down at low values since it ignores the exiting heating power which is also part of the "power out".

This fusion power amplifier sits at the heart of every fusion energy concept. What happens inside the vacuum vessel and the plasma, what actually increases or decreases Q_{\rm sci}, is the subject of a future article. For now, we treat the plasma as a black box with one knob: Q_{\rm sci}.

Every fusion energy company today is building a machine to create this power amplifier. The goal is to place it inside a system that converts the output power into electricity, feed a portion back to sustain the heating, and sell the surplus. The question is: how much Q_{\rm sci} do you need to have surplus to sell?

Converting power exiting to electricity

Figure 2 shows the simplest version of this idea. The power exiting the vacuum vessel (both fusion power and the heating power that also exits) enters a conversion step that produces electricity.⁸ Some of that electricity recirculates back to power the heating. If there's surplus "net electricity" it's sold to the grid. If not, power is drawn from the electrical grid.

The new control here is the efficiency of converting the total power exiting to electricity. No conversion process is perfect. A conventional thermal cycle (steam turbine) converts roughly 30% to 40% of the thermal energy to electricity, with the rest going to waste heat (labeled here as "rejected heat").

Try setting the efficiency to 33% and dragging Q_{\rm sci} down. You'll find that with this assumption, below Q_{\rm sci} = 2, no electricity reaches the grid at all.⁹ Every watt goes back into heating. Then try increasing the conversion efficiency. You'll see that at high conversion efficiency, net electricity can be achieved at a lower Q_{\rm sci}, even less than one! At high efficiency, most of the heating energy is recovered and recirculated, so even a modest amount of fusion power can reach the grid.

This is a fundamental tradeoff: lower conversion efficiency demands higher Q_{\rm sci}. More efficient conversion means a lower Q_{\rm sci} threshold for providing surplus electricity to the grid.

This version of the power plant makes two optimistic assumptions: that converting recirculating electricity to heating power is perfectly efficient, and that no power is consumed by the plant's other systems. We'll address these issues in the following sections.

Converting electricity to plasma heating

Converting electricity into heating power, whether through neutral beams, microwaves, or laser light, is never 100% efficient.

Figure 3 adds the heating system efficiency. With the conversion to electricity slider at 33%, try setting the heating system efficiency to 10% (the efficiency of some modern lasers converting electricity to laser light). What Q_{\rm sci} is required to put electricity on the grid?

Then set the heating system efficiency to 60% (the efficiency of some radio wave heating schemes in magnetically confined systems) and keep the conversion to electricity slider at 33%. What's the minimum Q_{\rm sci} needed to put electricity on the grid in that case?

Steady state and pulsed operation

The diagrams above show time-averaged power flows, which is the natural picture for steady-state systems like long-pulse tokamaks and stellarators. It's also an appropriate picture for pulsed systems averaged over many pulses, but it's useful to see how the repetition rate affects that average.

Figure 4 includes two modes, steady state and pulsed, and lets you adjust the heating level directly. You can see heating power values in Watts (steady state) or heating energy per pulse in Joules (pulsed mode) and get a sense of the level of electrical power delivered to (or drawn from) the grid.

In pulsed mode, try simulating a laser fusion system that delivers 2 MJ heating pulses at 10% heating system efficiency and achieves Q_{\rm sci} = 100. To deliver continuous electrical power, have the system repeat once per second (1 Hz). What's the net electrical power of such a system, assuming 40% conversion to electricity efficiency? What's the net electrical power if you then double the repetition rate to 2 Hz?

Then in steady-state mode, try simulating a stellarator heated by 50 MW of 50% efficient neutral beams and achieving Q_{\rm sci} = 10. What's the net electrical power of such a system assuming the same 40% conversion to electricity efficiency?

Explore further

We now have a reasonably complete picture of the basic power flows in fusion power plants. You're invited to explore the full simulator, which includes the house load mentioned earlier, split power conversion streams applicable to different fusion fuels, and blanket energy multiplication.¹⁰ Share design points of planned fusion machines and other interesting configurations on social media and embed them in other websites using the share and embed buttons on the full simulator.

The next article in this series will look inside the plasma and answer the question: what determines Q_{\rm sci}?

Sign up to our newsletter below to be alerted when it's released!

Questions? Email us.