Pulsed Power

^{K. Huppert}

Pulsed Power in Space: Why Peak Power Matters More Than Average Power

Spacecraft are chronically starved of power. A typical communications satellite might sustain a few kilowatts from its solar arrays. The International Space Station, with its acre of panels, generates roughly 100 kW. Even the forthcoming Kilopower fission reactor, a transformative technology for deep‑space missions, delivers only 1–10 kW of continuous electrical output[reference:0]. The Russian Zeus TEM nuclear tug, with its 500 kW reactor, represents the upper bound of near‑term space power[reference:1]. These are average power figures — the steady, round‑the‑clock output that a spacecraft can sustain.

Yet many of the most important things a spacecraft can do — ionise a propellant, compress a fusion target, fire a particle beam, or trigger a nuclear reaction — require power levels that exceed the spacecraft's average output by factors of a thousand or more. The resolution of this apparent contradiction lies in pulsed power: the art of storing energy slowly and releasing it quickly.

This article explains why pulsed power is not merely a convenience but a fundamental enabler of high‑performance space systems. It covers the physics of energy storage and fast discharge, the role of Linear Transformer Driver (LTD) capacitor bricks as the enabling technology, and the applications that become possible only when peak power is decoupled from average power.

The Fundamental Problem: Average Power Is Not Enough

Consider a spacecraft with a 10 kW Kilopower reactor — a substantial power source by current standards. If that spacecraft needs to fire a pulsed plasma thruster, the thruster might require a 100 MW pulse for 100 µs to ionise and accelerate a propellant slug. The instantaneous power demand is 10 000 times greater than the reactor's continuous output.

The spacecraft cannot draw that power directly from the reactor. It must accumulate energy over a much longer period — seconds, minutes, or even hours — store it in a compact reservoir, and then release it in a single, precisely timed burst. The reactor sees only the gentle, constant load of recharging the reservoir. The thruster sees the violent, brief spike it needs to function.

This is the essence of pulsed power in space:

(1)

\begin{align} P_{\text{peak}} = \frac{E_{\text{pulse}}}{\tau_{\text{pulse}}} \gg P_{\text{average}} = \frac{E_{\text{pulse}}}{T_{\text{recharge}}} \end{align}

where \(\tau_{\text{pulse}}\) is the discharge time (nanoseconds to microseconds) and \(T_{\text{recharge}}\) is the interval between pulses (milliseconds to seconds). The ratio \(T_{\text{recharge}} / \tau_{\text{pulse}}\) — the pulse compression ratio — can reach \(10^3\) to \(10^6\) or more. A system that averages 10 kW can deliver peak powers in the gigawatt range if the pulses are sufficiently brief and the storage sufficiently dense[reference:2].

Why Shorten the Delivery Time?

If the average power is the same regardless of pulse duration, why not deliver the energy more slowly? The answer is that many physical processes have a threshold behaviour: they simply do not occur at all below a certain power density.

Plasma breakdown. Ionising a neutral gas into a conducting plasma requires an electric field that exceeds the Paschen breakdown threshold. This field strength corresponds to a minimum instantaneous power density. A slow trickle of current, no matter how much total energy it carries, will never strike the arc.

Magnetic compression. Z‑pinch fusion and Mini‑Mag Orion concepts rely on rapidly rising currents to generate the immense magnetic pressure needed to compress fissile or fusion fuel to criticality. The magnetic pressure scales as \(B^2/2\mu_0\), where \(B\) is proportional to the instantaneous current. A current rising over seconds produces negligible compression; the same current rising in 100 ns produces gigapascals of pressure[reference:3].

Beam‑target interaction. In an accelerator‑driven nuclear system, the pellet must receive its driver energy before it disassembles. A typical pellet might expand and disperse on a timescale of tens of nanoseconds. If the beam power is spread over microseconds, the pellet has already blown itself apart before the energy arrives. Only a sufficiently fast pulse can couple energy into the target while it remains dense.

Efficiency of energy transfer. Many pulsed processes — inductive plasma acceleration, wakefield generation, ablation — exhibit a strong dependence on the rate of energy delivery. Faster pulses produce higher instantaneous fields, which drive more efficient coupling to the load before resistive losses and thermal conduction can bleed the energy away.

Thus the motivation for pulsed power is not merely to match the impedance of a load, but to reach the physical regime in which the desired process operates at all.

The Enabling Technology: LTD Capacitor Bricks

The workhorse of modern pulsed power is the Linear Transformer Driver (LTD) . An LTD is a modular, compact architecture in which many identical "bricks" — each containing capacitors and a switch — are arranged in parallel around a central magnetic core. When triggered, the bricks discharge simultaneously, their currents summing inductively to produce a single, fast‑rising output pulse[reference:4].

An LTD brick is deceptively simple. A representative university‑scale brick consists of two 40 nF capacitors charged to ±100 kV, connected in series and switched by a spark gap, delivering 25 kA to the load[reference:5]. Larger systems stack tens or hundreds of bricks into "cavities." The Sandia National Laboratories 118 GW LTD cavity uses 20–24 bricks, each containing two 80 nF capacitors, to generate a sub‑100 ns rise‑time pulse[reference:6].

The advantages of the LTD architecture are particularly relevant to space applications:

Modularity. Each brick is an independent, identical unit. A failed brick can be bypassed or replaced without affecting the rest of the system. This is essential for long‑duration missions where maintenance is impossible.

Compactness. Unlike traditional Marx‑generator/pulse‑forming‑line systems that require large oil or de‑ionised water tanks for insulation, LTDs are inherently compact[reference:7]. The bricks share a common ground potential, eliminating the need for high‑voltage isolation between stages[reference:8].

Fast rise time without external pulse forming. LTDs produce their characteristic fast rise time (~100 ns) intrinsically, without additional pulse‑sharpening stages[reference:9]. This reduces mass and complexity.

High repetition rate. Solid‑state LTDs using MOSFETs or IGBTs can operate at repetition rates of 50 Hz or higher, producing a train of identical pulses rather than a single shot[reference:10]. This is critical for propulsion applications that require sustained thrust.

Scalability. Voltage is added by stacking stages; current is added by adding bricks in parallel within each stage. The architecture scales linearly with the desired output.

The transition from spark‑gap‑switched LTDs to all‑solid‑state LTDs (SSLTDs) using silicon‑carbide MOSFETs or series‑parallel IGBT arrays has been transformative for space applications. SSLTDs eliminate electrode erosion, reduce jitter to sub‑nanosecond levels, and enable the high repetition rates needed for continuous thruster operation[reference:11][reference:12]. A recent 30‑stage SSLTD prototype achieved 279 kV, 3.1 kA output with 77 ns pulse width at 50 Hz repetition rate — a peak power of 0.9 GW from a table‑top system[reference:13].

Pulsed Power in Space Propulsion

The most immediate application of pulsed power in space is electric propulsion. Several thruster classes are fundamentally pulsed devices that cannot function without high peak power:

Pulsed Plasma Thrusters (PPTs). The simplest and most flight‑proven pulsed thruster. A capacitor is charged to several kilovolts, then discharged across a solid propellant (typically PTFE/Teflon). The arc ablates and ionises a tiny quantity of propellant, which is accelerated by the self‑induced Lorentz force. PPTs have flown on dozens of spacecraft since the 1960s. Their peak power during the microsecond‑duration discharge can reach megawatts, while the average power drawn from the bus is a few watts[reference:14].

Pulsed Inductive Thrusters (PITs) and FARAD. In these electrodeless thrusters, energy is stored in a capacitor bank and discharged through a flat inductive coil. The rapidly rising magnetic field ionises a gas puff and induces an azimuthal current in the resulting plasma, which then couples to the radial magnetic field to produce axial thrust via the Lorentz force[reference:15]. The Faraday Accelerator with RF‑Assisted Discharge (FARAD) separates the pre‑ionisation and acceleration stages, allowing independent optimisation. These thrusters require peak currents of tens to hundreds of kiloamperes with sub‑microsecond rise times — precisely the regime where LTDs excel.

Metal Plasma Thrusters. A newer class of pulsed thruster that uses solid metal propellant (e.g., aluminium, titanium) instead of gas or Teflon. A high‑voltage pulse strikes an arc between the metal cathode and an anode, vaporising and ionising a small quantity of the cathode material. The resulting plasma jet exits at velocities of 15–20 km/s. These devices are compact, require no tanks or valves, and can operate on inert, non‑hazardous propellant that can be sourced from in‑situ materials[reference:16].

Nuclear Propulsion. At the high‑power extreme, pulsed power enables the compression and ignition of nuclear fuel. Mini‑Mag Orion uses a Z‑pinch driven by a massive pulsed power system to compress fissile material to criticality, achieving specific impulses above 10 000 s with thrust‑to‑weight ratios of 0.2–10[reference:17]. Inertial confinement fusion concepts for propulsion similarly rely on pulsed power to drive the compression.

Beyond Propulsion: The Wider Role of Pulsed Power

Pulsed power is not confined to the engine. It permeates the entire spacecraft architecture wherever a process demands a threshold power density:

Communications. Deep‑space optical communication lasers require high peak power for pulse‑position modulation. The laser diode may produce watts average but kilowatts peak during the sub‑nanosecond pulses that carry the data. A compact LTD‑based pulser can drive the laser directly.

Scientific instruments. Lidar, ground‑penetrating radar, and plasma wave experiments all benefit from high peak‑power pulses. The resolution of a radar is proportional to its bandwidth, and bandwidth demands short pulses.

In‑situ resource utilisation. Drilling, fracturing, and material processing on asteroid or lunar surfaces may use pulsed power to generate shock waves or plasma discharges that break rock more efficiently than mechanical methods. The power supply on a small lander cannot sustain continuous high power, but it can charge a capacitor bank for periodic pulses.

Attitude control. Pulsed micro‑thrusters used for fine pointing and desaturation of reaction wheels operate in the milliwatt average regime but require kilowatt‑scale peak pulses to fire reliably.

The Thermal Paradox of Pulsed Systems

A common objection to pulsed power in space is that the waste heat generated during the pulse must be rejected by radiators sized for peak power, negating the mass advantage. This is a misunderstanding.

In a properly designed pulsed power system, the energy storage elements (capacitors, inductors) are charged slowly, and the waste heat from resistive losses is distributed across the entire charging interval. During the brief discharge, the energy flows into the load — the propellant, the plasma, the beam — not into the spacecraft structure. The radiators see only the average power loss, which is dominated by the charging efficiency, not the peak discharge power. The spacecraft thermal design is therefore driven by \(P_{\text{average}}\), not \(P_{\text{peak}}\)[reference:18].

This is a critical insight: pulsed power decouples the thermal load from the peak electrical load. A spacecraft that fires 1 GW pulses at 100 Hz with 0.1 % duty cycle dissipates only the losses associated with its 100 kW average power throughput. The radiators are sized for 100 kW, not 1 GW.

The Pulsed Power Philosophy for Space Engineers

The iSpaceE engineer approaches power management with a clear mental model:

Identify the threshold. Every physical process has a minimum power density below which it does not function. Determine that threshold before sizing the power system. Do not assume that a process can be performed "slowly" — many cannot.

Match the storage to the load. Capacitors store energy in electric fields and release it in nanoseconds to microseconds. Inductors store energy in magnetic fields and release it in microseconds to milliseconds. Batteries store energy chemically and release it in seconds to hours. Choose the storage technology that matches the pulse width required by the load.

Design for the average, switch for the peak. The primary power source — solar array, Kilopower reactor, YaEDU — need only supply the average power. The pulsed power subsystem converts that steady flow into the peak pulses the load demands. This is the architectural principle that makes high‑power space systems feasible on limited prime power.

Exploit modularity. LTD bricks are inherently modular and fault‑tolerant. A space‑rated LTD should be designed so that individual brick failures do not compromise the pulse train. Redundancy is built into the architecture, not added as an afterthought.

Respect the thermal rhythm. The radiators respond to average power, but the instantaneous heating of switches, capacitors, and magnetic cores during the pulse must be managed by local thermal mass and phase‑change materials. The thermal design of a pulsed power system is a problem in transient heat conduction, not steady‑state rejection.

iSpaceE Academy

International Space Engineering

Public Pages

For Registered Interest Only