Backend
Techno-Economic Boundary Analysis of Small Modular Reactor Cogeneration for Hyperscale Data Center IT and Cooling Loads
arXiv eess.SY0 views
arXiv:2608.10999v2 Announce Type: replace Abstract: Growing data-center electricity demand has increased interest in colocation with small modular reactors, but the conditions under which reactor power and steam-driven cooling reduce energy costs remain uncertain. Cooling can account for 20-40% of facility electricity use, making steam extraction a potential source of savings as well as a loss of turbine generation. We co-optimize hourly reactor dispatch, steam extraction, absorption cooling and grid exchange for a 200 MW$_\mathrm{e}$ data center in the Electric Reliability Council of Texas (ERCOT) region, using a 109-entry scenario set spanning capital, market, policy, financing and cooling efficiency. At 2023 mid-range reactor capital, the nuclear configurations cost 54.9-67.7% more than grid supply even with the Section 45Y production tax credit. With the other 2023 assumptions held fixed, cogeneration reaches grid parity near \$5000 kW$_\mathrm{e}^{-1}$. At nth-of-a-kind capital, nuclear supply is 85.9-98.7% cheaper than grid procurement in 2023, but between parity and 34.8% more expensive in the low-price 2024 market. A carbon price of \$53.1-64.0 tCO$_2^{-1}$ closes the mid-range gap when exports receive credits at the hourly average grid emission factor; the resulting negative balances are attributional and do not represent physical carbon removal. Absorption cooling responds to hourly electricity prices and supplies 32.6% of annual cooling, but adds \$9.7 million yr$^{-1}$ relative to the reactor-only plant. Absorption breaks even at an installed cost of \$26 kW$_\mathrm{c}^{-1}$ at baseline efficiency and \$529 kW$_\mathrm{c}^{-1}$ on a legacy-efficiency campus, compared with surveyed commercial prices of \$450-1200 kW$_\mathrm{c}^{-1}$.
Source:https://arxiv.org/abs/2608.10999PreviousThe Intrinsic Inertia, Droop, and Nonlinear Stability of Observer-Synchronized Grid-Forming ControlNextGraph-Native Attention Acceleration for Attack Detection in Cyber-Physical Systems
Related posts
What Jobs Can a Fresher Get After Learning Python? A Practical Guide
0dev.to
Defensive Coding for Real-Time Scoreboard Components
0dev.to
Dev News Digest: 19 Nov 2026, 06:00
0dev.to News
Download Website as HTML: Get a Runnable Project From a Figma Capture
0dev.to Design
A bilingual article’s canonical URL changes from a draft decision to an API result
0dev.to Discuss
Signs of Wear: Evaluating Intelligent Speed Assistance Under Synthetic Adverse Conditions
0arXiv eess.IV