Dear reviewer, thank you for the comment. Regarding the scalability, the model inputs allow for any size and number of turbines as the majority of costs are per unit of installed power. As such, different farm capacities can be replicated and studied, with economy of scale effects also applicable for larger farms. Each scenario was fixed at 1GW in this study for consistency. For reproducibility, the model uses metocean conditions from different sites, and ERA5 data can be applied to any maritime area covered by its grid. relying solely on new site-specific metocean data.

Regarding uncertainties and limitations, specific spacing and layout of the WTGs are not considered in the modelling. We felt this is not the focus of the study, for either farm layout for O&M or energy production. Standard spacing of around 1km per WTG was used. Energy yield does not consider layout and complex effects like blockage and wake effects. Although each case is taken as a separate entity, it should not be assumed that it is the only operational zone in the total available area, especially considering the 2GW currently available for auction. Bathymetry is considered as the depth changes significantly along the Portuguese coast, but seabed conditions are not considered for complex foundation costs or an optimised export cable route. Onshore cable route, substation and grid connection location are not considered in the model again due to added complexity, with simplified grid connection costs only up to the onshore substation. There are many reasons for omitting these costs, such as the variability of connections depending on farm, offshore layout and connections, shared connections, and existing onshore grid infrastructure.

For O&M and installation limitations, costs are top-down costs and do not consider the specifics regarding vessels, weather windows, logistics strategy and contractual arrangements. They do however consider the site severity and are distance-based cost functions. Port distances were measured using straight-line distances to port which may be slightly different in practice. Cable routes were also simplified and are mainly straight-line cost functions from site to shore. A summary of these limitations for both CAPEX and OPEX will be included within the “future work section, or as a specific limitations section to the model section.

Regarding previous work in this field, parts have been modelled, for example LCoE mapping in Portugal (Martinez, A., & Iglesias, G. (2022)). This study examined all of the European coastline, and as such the detail to the Portuguese area was less granular than our study due to the scale of the study. The economic feasibility (Castro-Santos, Laura, et al., 2020) examined LCoE in bands rather than specific development areas, with influence from wind and bathymetries. Also, tariffs were based on two fixed feed-in-tariffs and did not consider the range of CfD support mechanisms we analyse. All metrics considered in this work were considered in theirs, such as LCoE, NPV, and IRR. The wording in the literature review will be tidied up to make these aspects more clear.

One assumption is that the cost data on the wind farm CAPEX the reference farm used in the literature for this study, and is linked in terms of scale economies and construction methods. The majority of costs come from two sources: BVG’s Guide to a Floating , and a report on floating wind site-based and metocean influenced O&M, “A Spatial-Economic Cost Reduction Pathway Analysis for U.S. Offshore Wind Energy Development from 2015–2030” by (Beiter et al, 2016). On costs, the BVG report is widely referenced and is useful due to being recently updated and based on a large-scale wind farm. The assumptions are therefore that scale effects for the report-based 500MW farm scale well to 1GW, and that the costs for a UK floating farm are transferable to the Portuguese cases used in this study. To these assumptions, Portugal has been used as construction sites for past floating farms and has good capabilities here.

On O&M, an assumption is that sites fit into the reported division of “mild, moderate, severe” sites which can make edge cases difficult, as stated in the previous response using the report by (Beiter et al, 2016). For simplicity in modelling, we kept these three and most sites were grouped into the severe category due to the exposed and often strong metocean conditions that Western Portugal experiences. Also, costs were retrieved from (Beiter et al, 2016) and all floating sites were off the coast of California, so these costs assume similar OPEX between North America and Portugal.

Other notable assumptions were the discount rate set at a base 8%, which is a broad value for financing costs but is well documented for early-stage floating wind. Another assumption relates to the depth profile, assumed to correspond to the bathymetry at the central point of each of the 36 grid squares representing the sites.

For the CfDs, modelling was based on reference availability for capability-based CfD schemes. To estimate future values, synthetic datasets for prices and wind speeds were generated, avoiding simple replication of historical data. Uncertainty was constrained by statistically matching the synthetic data tohistorical distributions as closely as possible. These CfD methods are experimental but well documented, and are expected to feature in future offshore wind support mechanisms. These assumptions alongside references to all cost sources will be stated in revised version of the paper.

I agree with this comment on the conclusion and the need to link the conclusions to a broader context. An additional benefit of floating offshore wind is its higher economic value in terms of GDP generated per unit of CAPEX, a ratio that has been shown to be particularly high in other European markets. This potential is further strengthened by the possibility of floater fabrication in Portugal, which has already produced large-scale units for the WindFloat Atlantic project. Floating offshore wind may also create added value for aquaculture and, in the medium to long term, for the fishing industry, due to the creation of protected areas around the installations. These aspects will be further developed in the revised manuscript.

              Regarding comparisons with other studies, limited comparisons can be made with fixed-bottom offshore wind due to the deep water off the coast of Portugal being generally unsuitable in almost all cases,

For comparisons with other geographies, this is a very interesting point and is not covered in great detail. (Myhr et al, 2014) look into LCoE of floating offshore wind also compared the main platform types with fixed monopile and jacket foundations. One of many interesting points from that paper is the dramatic rise in costs as water depth increases. At the time of study, the limit for fixed foundation was 50m which is not being challenged, so reference to this and future options for the Portuguese market (could fixed ever be possible?) should be considered.

              A study on wave energy (Castro-Santos, 2015) reported LCoE values of approximately 95-180€/MWh at a discount rate of 7.5%. Applying similar discount-rate sensitivities to current conditions yields an estimated range of 126-239€/MWh today. This would place wave energy in a more favourable position than floating offshore wind in Portugal; however, this comparison does not account for changes in CAPEX since the time of the study.

              A forecast of energy and subsequent LCoE for all of the European coastline (Thomas, 2026) found that floating offshore wind LCoE in Portugal was around 100-180€/MWh, lower on average but showing good agreement with the best performing sites analysed in the current study. (Thomas, 2025) examined floating wind in Spain, with 120€/MWh for the site just above the Portuguese border on the Western Coast. (Castro-Santos, 2020) in the “Economic feasibility of floating offshore wind farms in Portugal” had higher values, including 304€/MWh for the most relevant semi-submersible platform.  The broader European coastline study by (Martinez and Iglesias, 2022) found a range of LCoE which showed detailed mapping and distinction off the Portuguese coast, 160€/MWh closer to shore and 140€/MWh in the far shore areas, with the cost effect of deeper locations on mooring costs unknown. These LCoE results frame the outcomes of our study well, ranging from 146-319€/MWh with bathymetry extremely impactful site performance. The UK AR7 floating wind LCoE is also approximately 249€/MWh, which again agrees well with the outcome of this study, albeit for a much smaller pilot farm of around 200MW.

              The goal of this paper is to highlight under what conditions floating wind is viable in Portugal, in particular in terms of financial support mechanisms. It is hoped that this could strengthen the case for increased governmental support for floating offshore wind by first outlining the challenges and then highlighting the associated environmental, economic, socio-economic and marine-economy benefits.

Specific areas that need clarification:

Lines 39-41: “Notwithstanding the lower costs of onshore wind and solar, (Silva & Sareen, 2021) argue that public attitudes vary, with community opinions and engagement often overlooked or bypassed, which presents an opportunity for floating wind development.” I’m not sure what this means; how does it present an opportunity for floating wind development?

This could certainly be clearer: The paper argues that onshore renewable projects often face stronger public opposition due to land-use and environmental constraints, and because community concerns, particularly in areas close to proposed developments, may be insufficiently addressed. In contrast, floating offshore wind offers additional space for renewable deployment with fewer perceived land-use conflicts, no nearby human habitation, and potentially lower environmental impacts, particularly in terms of biodiversity. Moreover, following construction, these installations may even deliver net positive effects for marine ecosystems. This will be worded more clearly in the revised manuscript.

Lines 161-162: “This method was applied to twelve locations, each at three distances to shore, to provide a comprehensive feasibility study for the whole of Portugal.” This was a bit confusing; I would clarify that the method was applied to twelve regions, each at three distances to shore, for a total of thirty-six distinct locations.

I agree this could be worded more clearly sense and will be changed in the next version.

Lines 298-299: “Geographical features including Sintra peninsular off Lisbon and also in the Southwest, Sagres zones also benefit from improved but local windspeeds.” Confusing text.

This is indeed confusing. This section highlights the geographical features that enhance wind speeds in the area and, in turn, improve the site’s overall suitability and viability for development. This will be reworded.

Figure 4: Why is the LCoE so much higher in the further sites in Sagres West compared to Sagres South? It seems like they are following the same windspeed/depth patterns, but there could be some variation in the data that I'm not seeing in this figure.

The two sites lie at the edge of the continental shelf. The southern site has a water depth of 245 m, whilst the western site reaches 2811m, which largely explains the significant difference in LCoE between the two sites. This will be more clearly pointed out in the text.

Lines 353-354: Name the “selected few sites” viable under current economic conditions.

The sites viable under current economic conditions are shown in Fig. 9: Viana (near), Porto (near, mid), Aveiro (near, mid), Figueira (near, mid), Peniche (near, mid), Lisboa (near, mid), Setubal (near), Sines (near), Sagres W (near), Sagres S (near, mid), Lagos (near), Faro (near). These sites will be more clearly listed in the text. The majority of edge-case areas fall into the marginally feasible category (corresponding to the 8-10% range, where 8% is the assumed discount rate). This sentence will be reworded for clarity.

Line 375: Was a sensitivity analysis shown for IRR?

Yes. This aspect was considered but omitted due to similar results with NPV sensitivity. This will be re-examined in the next iteration as IRR is an often-preferred metric due to the fact that it can be compared with to the discount rate.  It is perhaps easier to understand to a wider audience which could improve the readability of this paper, and convey the message better.

The technical corrections will be addressed and the comments are appreciated.

The manuscript would benefit from a clearer explanation of what constitutes “economic viability” for FOW. Specific economic thresholds or benchmarks used to define viability would be of value. Also, providing guidance on reasonable ranges for financial metrics such as internal rate of return (IRR) and net present value (NPV) would help contextualize the economic analysis and improve interpretability for readers.

• Economic viability is essentially positive NPV or IRR>discount rate, but that should be conveyed within the Portuguese context, and also within the context of what a future energy system in Portugal looks like, how will demand increase to ensure prices are viable for increased levels of penetration. I think this is a broader point on the Energy system of Portugal as a whole, and will be considered as such.
• What we did to fix:
  • Added clearer definition of what is a economically viability considers (bottom of introduction)
  • Define economic viability into the future energy system of Portugal
  • Define reasonable ranges, from literature (Noothout, et al., 2016)
  • Give practical responses on the results, not just the values and a better discussion of results

A few aspects of the modeling framework would benefit additional clarification. First, a clearer explanation of the parameter sweep, and scenario structure would improve reproducibility and reader understanding, Referencing Table 1 explicitly in this context may be helpful. Second, the extrapolation methodology used to scale project capacity for the OPEX analysis could be described in greater detail to better understand how this was done. Third, assumptions around underlying the learning rate discount should be explicitly stated and justified. Forth, it would be helpful to include representative CAPEX values from the referenced BVG report directly in the manuscript for ease of reference. Lastly, specific assumption for the system losses may be important to report explicitly since the energy generation of the projects are such as significant drive of LCOE reductions.

• I agree and I think the framework could be better explained in the text and also in Fig.1. The table should be referenced better and improved to give more useful information on the farm, and remove unnecessary values and repetition.
• On the second point, yes the OPEX method could be improved, the raw values may not add much here and so could be simplified and referenced to the original NREL work.
• Learning rate can be backed up well for early-scale commercial floating wind and will be better defined, alongside representative CAPEX. On losses, they are very general as the main focus is on the high-level assessment and changes to CfD, but these can certainly be defined better even though they are equal for all sites, we did not run wake analysis either at intra-farm or inter-farm level due ot processing time and the focus on this study
• What we did to fix:
  • Grouped pie charts for base CAPEX in Fig.2
  • Table 2 now includes base CAPEX values and low/high range for sensitivity
  • Improved Table1 and referenced parameters within text
  • Improve OPEX calculations Eq5 with improved explanation
  • Define specific losses better in text, in Table 1.

Given that different floating platform technologies may exhibit differences in material costs and impact variances in operational and maintenance costs, these distinctions seem important and should be addressed explicitly. Similarly, it would be useful to clarify whether site selection considers the three different types of FOW substructures.

• We chose the semi-submersible due to similar technology already present in Portugal and already developed supply chains here for such structures. Also, the general baseline CAPEX assumptions from the BVG study also use the steel semi-submersible as the chosen technology and it was logical to remain with this. A comparison of the three structures would be useful but then would add a complexity and different cost source. An interesting idea and one which was considered for future work.
• What we did to fix:
  • Better define the choice of SS platform, reinforcing the choice of semi-submersible across all distances and depths
  • Clarify the decision of SS in the text and method, why was it chosen as the base case and not others.

A few technical and economic terms would benefit clearer definition or expanded explanation, a few including strike price, spot price, reference volume, reference generator, windfall event, restrictive phases, learning rate discounts. Also, acronyms such as “TLP” may be better spelled out to ensure accessibility to a broader audience.

• That is understandable and will be updated, the paper combines technical and economic analysis, and such can speak to two audiences and key terms should be better defined
• What we did to fix:
  • Explained acronyms and technical and economic terms which may be unclear to a wider audience
  • Removed the statement on reference volumes for the traditional CfD, as this was incorrect (based on farm generation, not reference generation).

Initial presentation of the LCOE equation with the time index term “t-3” could use an explanation when describing the equation. Another review of the authors to ensure accompanying explanation for all equations and variable are defined would be worth the time. Additional context for how financial inputs, including IRR targets, discount rates, and cost escalation assumptions are selected would also strengthen transparency and understanding.

• This was only done as CAPEX is split into the four years before the start of operations (t=-3, … , 0). This can be defined better, and so can the other equations perhaps with more practical context. I believe most terms in suitable equations are added but if space allows in the manuscript these will be further elaborated. Cost escalations tend to be linear and already mature in terms of economies of scale but more explanation will be added if required
• What we did to fix:
  • Defined the economic assessment indicators and their practical significance and reason for selection
  • Reevaluate equations and their terms and explanations, LCoE equation t=-3 justification added
  • Add a table of economic variables, terms and reason for selection

Addition clarify around a couple of interpretive points would strengthen the manuscript. Statements around hydro-battery investments, community engagement imply negative sentiment toward onshore wind and solar, and whether observed variability in design parameters reflects a lack of industry standardization or is primarily driven by site-specific engineering requirements.

• This is an important and expansive point to the wider question of marine renewable and renewable energy integration in Portugal. Sections added to confirm that floating wind will not develop without adequate grid and demand to avoid low prices.
• The study on solar was very locally-focused in its surveys, and the broader acceptance of renewables has been added to the manuscript.
• Further points on the design standardisation have been added, with both of importance for platform cost reductions and overall design optimisation. A key study limitation is added here, keeping all platforms equal irrespective of site conditions, albeit most are classed as “severe” in the O&M analysis except some Algarve sites in the Southeast.
• What we did to fix:
  • Added the need for faster demand-side response (line 41)
  • Elaborated on the social acceptance around line 45
  • Added further comments on design standardisation, with the paper in question from 2014 which would contribute to the high variability.

A couple of minor comments for consideration. First, additional comparison between competitive sites and what parameters are driving the appeal. Second, mention of specific cost reduction strategies. Third, expand descriptions where key technical or economic assumptions are introduced. Forth, strengthen explanatory text for clarity on equations. Finally, consider integrating key data points from referenced reports directly into the text to improve readability and reduce reliance on external sources.

• I agree that more statistical and qualitative comparisons could drive insights, including the changes in depth, wind speeds and general metocean conditions. The cost reduction strategies play more into future design optimisation but also rely heavily on the future economic context, this will be considered where possible. I believe the comment on the clarity of equations has already been mentioned and will be actioned. On the reports, integration into more tables will be considered.
• What we did to fix:
  • More insights and discussion on the comparison between the well-performing floating wind areas, and further add description and clarity into economic indicators and equations (already listed as action).
  • Cost reduction strategies (possible future developments and consider new technologies away from semi-subs) and any broad consideration of the future for floating wind capex from new reports.