Hybrid Energy Systems Sizing for the Colombian Context: A Genetic Algorithm and Particle Swarm Optimization Approach

Round 1

Reviewer 1 Report

The article deals with the optimal sizing of an Hybrid Renewable Energy System, with an application to the Colombian context. The number and type of wind turbines and solar panels are considered as optimization variables as well as the number of batteries in the storage system. A Genetic and a Particle Swarm Optimization algorithms are used and compared to solve the problem. the Loss Power Supply Probability (LPSP), the Total Annual Cost (TAC) and the Levelized Cost of Energy (LCOE) are considered in mono or multi-objective formulations.

The paper is globally well written, the simulation tests are performed using an interesting and relevant dataset, and the results are analysed.

Here are some comments that should be addressed before final publication:

• Considering the model in section 2, it appears that the consumption should be entirely fulfilled from renewable energies (and storage). A mixed system with classical production units may be more interesting to obtain an efficient energy system.
• (Eq. 7) and (Eq. 8) are not in accordance with (Eq. 15). Or do we consider that all renewable energy units have their own storage system? In the optimization part, it seems that we have a single global storage system?
• (Eq. 26) considers that all consumers have exactly the same water consumption curve?
• The optimization variables can only take some integer values. I do not see how this property is taken into account in the PSO algorithm, whose presentation in the paper, and especially (Eq. 33) is dedicated to continuous optimization problems.
• The multiobjective formulation with weights equal to 0.5 could have been extended to a real multiobjective solution, leading to the Pareto front that ca be obtained considering other combinations of the criterions.
• A comparison is done with the formulations. But, obviously, when we want to obtain the lowest value of a criterion, the methods, that are actually the best to do so, are the ones that explicitly take into account this criterion in the problem expression.
• I do not know if the comparison of the solutions obtained with PSO and GA, are completely fair. Of course, with the solutions that have been found, we can give some arguments about their quality. But, these solutions strongly depend on the particular run of the algorithms and the way they have been tuned and implemented.

Author Response

Please find attached file with answers to the reviewer comments.

Author Response File: Author Response.pdf

Reviewer 2 Report

Because the power produced from renewable energy sources is characterized by its variability, a solution to meet the demand for a load from renewable sources is to combine several energy systems together and form a hybrid system. The most important for a hybrid energy system is that the system should be energy and economically atractive and therefore the system should be optimally sized. To design and assess the hybrid energy system performance is a complex task duet o the randomized nature of alternative energy sources, electrical load profile, as well as the non-linear response of system components. Various methods have been proposed to solving this problem, some based on more traditional approaches and some based on heuristic approaches such as the Genetic Algorithm and Particle Swarm Optimization used in this paper.

The paper is well written and well described and easy to read. The methods used are appropriate and the conclusions are consistent with the obtained results. The findings are relevant and interesting to the readers of the journal.

One minor comment: Page 8, Table 1. The parameter e_c,gb should have units of measure [kWh/m³] not in [kW-h/m⁴].

Author Response

Please find attached file with answers to the reviewer comments.

Author Response File: Author Response.pdf