Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

• The paper resolves data-center placement with energy constraints in an ILP with binary variables. While timely (e.g., "East Data, West Computing"), the question raised in Section 1 is worded too vaguely and would be enhanced by closer specifying the gap between optimization employed herein and existing models of siting (e.g., p-median, MCLP variants). Clarify exactly what is innovative beyond incorporating a standard energy cap.
• Connected work (Section 2) is broad but somewhat catalog-based. MCLP, p-median, and ensemble coverage models, for example, are all mentioned, yet the paper does not correlate these to your final constraints (Eqs. 12–17) or show which features (coverage, latency, reliability) are maintained or relaxed within your model. Add a comparative table comparing earlier assumptions to your model.
• Decision variables are established (x, y, z for siting, assignment, interconnection) and then employed in constraints (Eqs. 12–15), but notation is not consistent (subscripts/indices occasionally trimmed). Normalize notation, clearly define index sets, and make constraints like interconnection (Eq. 15) network-feasibility-justified.
• Assumptions in Section 4.1 are based on true locations (12 demand cities; 8 candidate sites) and parameters from "China Power Intelligence Network" but no specific citations, dates, or datasets provided. Provide a data-availability statement and a supplemental file with the exact numeric inputs (Tables 1–4 fully completed) to allow replication.
• The contention that energy constraints increase facility count (from two to three) is valid (Tables 5–6), but it needs a justification of feasibility. For example, which constraints (Eqs. 16–17) constrain at optimality? Supply dual values or at least which facility site capacity limits (R_ls, S_ls) become effective to justify the additional site.
• The conclusions are that "cost is lowest in low-electricity-price areas," as would be anticipated. Enhance scientific contribution by illustrating a trade-off frontier (cost vs. energy consumption vs. latency), potentially through ε-constraint or weighted multi-objective runs, rather than just two discrete cases.
• Reproducibility: Solver used is MATLAB YALMIP + CPLEX. Provide solver parameters (gap tolerance, time limit), model size (|R|, |S|, number of binaries/constraints), and computing time. A reference to a GitHub repository with model files and data (anonymized if needed) is highly encouraged to meet open-science standards.
• References are 2020–2024 sources and sectoral work but a few entries appear to be a duplication or in inconsistent format (e.g., Turek & Radgen duplicated twice) and some journal details appear to be mismatched. Check consistency (publisher, volume/issue, page numbers, DOIs) and in Energies style.

Comments on the Quality of English Language

The paper is legible but needs editing for application to articles, consistency of verb tense, and technical language (such as repeated "as shown," inconsistent capitalization, and omitted words/indices within equations from time to time). There are a few long sentences with nested clauses containing hidden meaning. Professional editing of the language for clarity, consistency, and readability is recommended.

Author Response

Please refer to the attached document for detailed responses

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors propose a layout optimization model based on energy consumption constraints that combines integer linear programming with binary decision variables.

Questions:

1. Are data centers located in transmission or distribution systems?

2. The proposed optimization model is multi-objective. How do the authors ensure that all functions are minimized during simulations?

3. Table 3 is broken across two pages. Avoid breaking tables.

4. The authors use the YALMIP toolbox and the CPLEX solver in Matlab to solve the optimization model. Every time this solver is applied to this optimization model, does it always converge to the same solution? Or might the results be different for each simulation? It would be interesting to present a statistical analysis for each simulation of the optimization model with this solver to evaluate the convergence results.

5. The authors stated that they included energy costs in the optimization model. How did adding this constraint affect the traditional data center allocation model? Are there clear benefits? The authors could conduct an analysis of the results.

6. Which parameters of the proposed optimization model can be defined by the user? Furthermore, what values ??of these parameters were used during these simulations?

7. The authors do not present a comparative analysis with other methods available in the literature. Comparative analyses allow us to assess the benefits and weaknesses of the proposed method.

8. The authors consider latitude and longitude values ??in the data center allocation model. But are these values ??sufficient? Wouldn't a power grid topology with information on impedance, hourly power flow, and current level be necessary for adequate data center allocation?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The article is devoted to the issue of optimizing the layout of data centers. The topic is very relevant due to the huge demand for such centers. The authors propose a new layout optimization model based on energy consumption constraints, which combines integer linear programming with binary decision variables. Of practical importance is the consideration of energy efficiency, availability of renewable resources and the features of the balance of economic benefits and environmental impact in the proposed model. The authors have experimentally proven that the energy efficiency scenario consistently reduces costs, however, energy constraints can lead to an increase in the number of required data centers. At the same time, there are a number of questions to the article:

1. The captions of Figures 4 and 7, as well as Table 3, are shifted.

2. Table 1 does not contain units of measurement for electricity prices.

3. The article does not explain how the availability of renewable sources is related to the optimization of the placement of data centers. Why are other sources not suitable for this task?

4. It is unclear which optimization method is used? What optimality criteria are proposed to be used?

5. In addition to the Cost per Unit Length of Optical Fiber parameter, shouldn't you also take into account the power consumed by communication devices?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

All previous comments have been addressed, improving clarity and rigor. The methodology and conclusions are now well-structured and aligned. No further comments.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors propose a layout optimization model based on energy consumption constraints that combines integer linear programming with binary decision variables.

The article has been improved, the contribution is good and all questions have been effectively answered.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have responded to all comments in full. There are no more questions.