Power-Delivery Topologies for a D–³He Tandem-Mirror Data-Center Burner: Solid-State-Transformer versus Direct-DC Conversion, Grid-Forming Control, and DC-Bus Fault-Ride-Through
Year: 2026
Type: dataset
Abstract: The MetroVolt burner takes a deuterium–helium-3 (D–^3He) tandem mirror and delivers its output as electricity to a co-located, high-density computing load. Because the burner's direct-energy converter (DEC) decelerates a charged-particle stream against graded collectors, it is a high-voltage DC source by construction, and the power-delivery architecture becomes a set of down-select decisions that a conventional AC plant never faces. We pose those decisions as formal optimisation and stability problems on a DC-native bus, derive their governing equations, solve them with a GPU-accelerated reduced-order campaign, and settle four of them. (i) Isolation and step. A cascaded source-to-load efficiency model shows the solid-state transformer (SST) is redundant: a DEC-direct path reaches 88.39% against 85.43% for an SST-fed path (+2.96 points with one fewer conversion stage), because the galvanic isolation an SST would provide collapses into a single silicon-carbide (SiC) dual-active-bridge (DAB). A conduction/switching/magnetic loss decomposition gives an arithmetic–geometric-mean optimum at f^=15.81 kHz and η_DAB=98.85%. (ii) Grid interface. A single-bus swing-equation model of a 6% generation-loss event on a 1 GVA island (H=3 s) shows the DC-native buffer makes the plant grid-forming: a sub-10 ms fast-frequency response lifts the frequency nadir by 0.527 Hz (49.37349.900 Hz, clear of under-frequency load-shed) and enables black-start and islanding a grid-following plant cannot. (iii) Fault protection. A closed-form LC-discharge analysis of a 10 kV bus (5 mF, 5 , Z_c=0.032 ) gives a 316.2 kA undamped resonant peak at 248 ; a 2 SiC solid-state circuit breaker bounds the fault to ~4 kA (1.3% of the peak), and the DEC's intrinsic current limit caps a source-side fault to ~3 kA against ~50 kA from the grid (17× lower). (iv) Ride-through. Cross-timescale energy sizing shows a full-IT bridge needs a ~12.5 t battery, so graceful load-shed to a ~2.5 t critical set, with a ~3.3 t black-start battery, is the design. The four decisions compose into one coherent architecture: DC-native throughout, simpler, lower in prospective fault current, and islandable, carrying the burner's frozen operating point (Q_E=1.318, direct-conversion fraction 0.49) to the load. Only technical metrics are reported; economics are out of scope. Key results (frozen anchors): eta_DAB = 98.85 %; kg^-1 = 7.2 ×10^5; DAB = 98.85 %; SST = 85.43 %,; DEC = 88.39 %,; N = 1 eq. Related: Paper page · De-risking register · 3D model · Learn more about Kronos Part of the 2026 Kronos publication series; independently re-run and stamped in the Kronos de-risking register (DOI 10.5281/zenodo.22645689). All numerical values are frozen design-point anchors; see the register. Public research artifact. No proprietary, financial, or supply-chain information is included. Part of the Kronos Fusion Energy 2026 Publication Series — 115 papers on a compact spherical-tokamak breeder (Hyperion) and D–3He tandem-mirror burners (Aegis / MetroVolt). Every numerical value is a frozen design-point anchor, independently re-run and stamped in the register. Zenodo record: 10.5281/zenodo.22645941 De-Risking Register (mothership): 10.5281/zenodo.22645689 Read online: https://www.kronosfusionenergy.com/publications/power-delivery-topologies-for-a-d-he-tandem-mirror-data.html Live validators: Physics Validation & Simulation © 2026 Kronos Fusion Energy — a family-owned American company. Licensed CC BY 4.0.
Source: Harvard Dataverse
DOI: 10.7910/dvn/xtvysq
Author: Ford, P I
Language: English
Topic: Advanced DC-DC Converters
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7212024196