Development of a Coupled TRNSYS-MATLAB Simulation Framework for Model Predictive Control of Integrated Electrical and Thermal Residential Renewable Energy System

Round 1

Reviewer 1 Report

A hybrid thermal-electrical renewable energy system in a Sonnenhaus standard 786 single-family house is investigated with a developed TRNSYS system model. The paper mainly focused on the description of the proposed approach.

In the reviewer opinion the paper quality is not sufficient for publication. 

The major concerns are reported in the following points:

• The introduction does not highlight the major contribution of the authors, which are the prons and cons of this approach with respect to other approaches developed in literature, and which is the novelty of the approach proposed.
• The paper is too long and the structure is not clear. Details of the practical implementation are provided while important information about the algorithm are briefly and not clearly explained. 
• Comparison against other approaches to show the advantages related to the use of the proposed one are not shown, the proposed approach aims at maximizing the RES exploitation and the selfconsumption, however other objectives in contrast with these are not considered at all by authors (costs, efficiency, etc) resulting in a poor results sections and limited conclusion.
• Information about the system sizing are not provided, several works in literature have shown how, sizing and management are strictly correlated in these kind of systems, authors should discuss about that.

Minor concerns are instead related to:

• Table 1- Primary Energy. What does this parameter means? Is the primary energy consumption in one year? The unit should be kWh/m2 y 
• line 400 - There is an error with the reference
• other several typos within the paper.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

1. It is better to change the title. The title of the article should not contain abbreviations.
2. Do not repeat words (phrases) from the title of the paper as keywords. The title and keywords have the same terms (Model Predictive Control; Renewable Energy, TRNSYS; MATLAB). It`s reduces the attractiveness of the article citation.
3. The purpose of the research should be emphasized more clearly.
4. Materials and Methods Section should be added.
5. Figures 2 and 3 has been pasted from the previous publication (Narayanan, M.; Mengedoht, G.; Commerell, W. Importance of buildings and their influence in control system: a simulation case study with different building standards from Germany. Int J Energy Environ Eng 2018, 842 9, 413–433, doi:10.1007/s40095-018-0281-9.). It is unacceptable.
6. Line 400: “Error! Reference source not found.” What does it mean?
7. Lines 510-511 (formulas 5 and 6): where are equations 1-5? All variables must be described.
8. Lines 526-547: Please, use units of measurement for variables.
9. Figures 10-14: why did the authors use two decimal digits after the decimal point if they are equal to zero everywhere?
10. Tables should be done according to instructions for authors.
11. Lines 904-905: “(accessed on Jan 13, 2017)”. It has to be done according to instructions for authors.
12. Line 910: “(accessed on Jun 8, 2018)”. It has to be done according to instructions for authors.
13. Line 968: “(accessed on Aug 14, 2020)”. It has to be done according to instructions for authors.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments for the authors:

1. Overall, 28 pages are too long for a paper. Please shorten the manuscript or create 2 paper instead.
2. According to my knowledge, “Sonnenhaus standard” is not a building energy standard, as Effizienzhäuser, Passivhaus, Niedrigstenergie- und Nuééenergie Haus, Effizienzhaus plus, Active House or Minergie-Standard. Please provide more precise information what a “Sonnenhaus” is.
3. Line 39: Please write the 2 sentence in 1.
4. Line 45: The word “eventually” is misleading. Increasing renewable utilization leads surely to primary energy consumption reduction.
5. Line 48: The word “the” is not needed
6. Line 48: Decentral sector is firstly meant in the manuscript and not evident what is the meaning.
7. Line 65: Please correct grammar in “uses” (use).
8. Line 100: HP is abbreviation for heat pump?
9. Line 136: Sentence starts with “And” – please avoid and revise sentence.
10. Fig.1: delete vertical black line at right side of fig.
11. Line 69: What is OFFIS abbreviation for?
12. Line 172-173: It is not clear whether the mentioned SFH is a real building or a simulation model.
13. Line 195: What is the demonstrator exactly? A real building with real HVAC system? Meteorological-Station?
14. Line 210-211: If a CHP is not reasonable in SHF, why then applied in this study?
15. Fig. 4: Correct the word “Definition” at Objective Function Definition. Please set this figure after the descriptive text section in the manuscript.
16. Line 400: Error, please provide reference.
17. Line 490: The reduction of the use of a secondary gas boiler system (for peak loads) is evident, still it reduces the investment-exploitation lower…
18. Line 647: The startup simulation is meant already multiple times in the manuscript. Please avoid doubling.
19. Line 662: Sentence starts with “And” – please avoid and revise sentence.
20. Line 712: Sentence starts with “And” – please avoid and revise sentence.
21. Line 783-784: In winter is still PV production available. Is it so low that winter operation optimization is not possible?
22. Conclusion section: Here, only a summary of the study provided. No new conclusions to be found. Please write here a real conclusion section, including new insights-conclusions.

Questions:

1. Why are the regulation time limits so high? (Table 2.) In the case of SOFC, I would definitely take it lower as it can be regulated flexibly. Is the of turn-on time 30 hours? If yes, you should always run at a minimum of 500W, which is not very realistic, it can generate a lot of losses to the system.
2. How optimal is a system and meets any standard, where the system consumes natural gas even in summer, even though the required electricity should be generated by solar panels (Figure 15)?
3. How can you guarantee that the new generations in GA remain feasible?

To summarize:

The paper handles a very interesting and innovative field in sustainable building technologies and optimization strategies, and, in addition, after consideration of my remarks and questions it can be recommended to publish.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

The article deals with a classical model predictive control (MPC) strategy to favour self-consumption using renewable energies in a family house, which is an interesting and current challenge. To get a more realistic description of thermal behavior of the house, the authors focused on white box model, i.e. physical model. In my opinion, the main contributions of the article concern the implementation aspects, with the coupling of a dynamic thermal simulation software (Trnsys) and an optimization problem solver (Matlab), and the comparison of three optimization algorithms (particle swarm optimization PSO, genetic algorithm GA and global pattern search GPS). Unfortunately, two limitations can be underlined for operational uses. Firstly, the weather and the demand forecasts are considered perfectly mastered. Hence, no uncertainty is considered herein. Secondly, the MPC  simulation time seems to be prohibitive for real applications, e.g. more than 6 hours of computation to get a control on 24 hours. Articles in the literature have already proposed much faster strategies using MPC or Stochastic Programming. Nevertheless, we can appreciate that the article is applied to a real house taking into account the full complexity of energy systems (PV, battery, fuel cell, gas boiler, etc). Before publication in the journal Energies, I recommend many minor revisions:

1) Please make a careful reading of the article and correct all the typos : replace “defenition" by “definition” in Figure 4, reference not found at line 400, replace “left” by “right” at line 722, replace “winter” by “spring” at line 761, …

2) Some phrasing can be confusing in the article. In agreement with the optimal control theory, “control variables” should be preferred to “manipulated variables”. Moreover, at line 397, loads are associated with “controlled variable” which is confusing for the reader, please find another phrasing for the loads.

3) In Section 4.3, I do not understand how the penalty terms are considered when solving the optimization problem. Are they added as additional terms into the objective function? If yes, are there weighted by coefficient? Please clarify this point.

4) After introducing Eq. 5, please add a sentence explaining that the notations are detailed at the end of section 4.3.

5) When reading the paragraph 499-510, especially the words “the same objective” (line 508) that the formulation of Eq. 5 is equivalent to the formulation of Eq. 6. I think this is not the case. If I’m correct, please replace “the same objective is written” by “another objective can be written” or “similar objective can be written”.

6) Usually, squared functions are considered to improve the convexity of the objective function involved in the optimization problem. And the increase of penalty is ensured by choosing bigger weight factors. I’m surprised by the choices of the authors at lines 515-517. Did you see these techniques, i.e. functions are squared to increase the penalty, in the literature? If yes, please add references.

7) I’m wondering if the “<” is correct in the first line of Eq. 9. I thought it was a “>”. Please verify this point. If there is a typo error, please correct all “<” and “>” in Eqs. 9 and 10.

8) Concerning the computation time in Section 5, please indicate the type of used computer with processor information (number of GHz). Precise if you made sequential calculations or parallelized.

9) In Figure 14, please add a graph with the room temperature for each algorithm and the room temperature setpoint.

10) How the proposed MPC strategy can be extended to take into account uncertainties (weather and demand forecasts) for practical uses? Please add some responses, for example, in the conclusion section.

11) Please add perspectives for the reduction of computation times such as the use of offline/online computations.

Author Response

Please see the attachment.

Author Response File: Author Response.docx