Optimal Data Center Placement for Distribution Networks Using Intelligent Algorithms

Amin Hajihasani and Mahmoud Modaresi present a techno-economic optimization framework to site data centers on distribution network buses by jointly selecting connection bus and DG size using an intelligent genetic algorithm.

• Main announcement/action: The paper develops a mixed-integer nonlinear optimization model and embeds an intelligent genetic algorithm in a multi-scenario decision framework with adaptive weight tuning to jointly select data center connection bus and size on-site DG (distributed generation). The converged case-study design selects bus 14 with 1.10 MW DG, reduces total losses from 202.67 kW to 129.37 kW, improves minimum bus voltage to 0.933 p.u., and reports a moderate investment cost of 1.33 MUSD.
• Background and details: The study uses the IEEE 33-bus radial system as a case study, formulates an objective combining active power losses, a voltage deviation index, and investment cost (location-dependent land price and unit DG cost), and explores three stakeholder scenarios (prioritizing losses, voltage quality, or techno-economic balance) plus automatically generated balanced scenarios until decision convergence. Submission: 26 Dec 2025; arXiv id arXiv:2512.21987; paper length 7 pages, 4 figures, 4 tables.