**4. Discussion**

The objective of this study was to question the dependence on fossil fuels in the future electricity system in Cuba. Isla de la Juventud provides an excellent case study for power system modeling due to its size and because it is an island system with similar characteristics to the Cuban main island power system.

The Cuban governmen<sup>t</sup> has applied a very stimulating strategy to deployment his policy to development RES within the country: (1) to open the foreign investment to achieve RES technologies and financial capital, even with 100% of the foreign investment, (2) to increase the interaction with international organizations in order to develop "absorptive capacity" on RES and technologies (3) to give autonomy for the local governmen<sup>t</sup> to decide their development strategy for increasing penetration with RES, (4) to develop a program to prepare the local governments to be involved with RES, (5) to involve universities with the local governments in order to enhance opportunities for different stakeholders to support the use of RES in local conditions, (6) to incentive private sector using RES. Consequently, the governmen<sup>t</sup> needs to identify and raise awareness of key policy questions and their implications for the long term development of the Cuban electric power systems means "to quantitatively calculate the direction of future energy policy and the implications of taking one pathway of energy sector development instead of others" [30].

It has been recognized in literature that desirable generation expansion scenarios or renewable energy targets depend mostly on policy priorities and on economic resources, rather than on technical grounds [15,16,18,31,32]. At the same time, the particular characteristics of a given isolated power system influences its capability to safely integrate increasing

shares of RES. Although this analysis is mainly based on planning alternatives, some technical and operational aspects of the system were considered for the selection of the scenarios. These include aspects such as generation capacity and technology, maximums and minimums of power by conventional generators, ramp rate, load share, generation dispatch, grid congestion and stability issues.

The system is made up of fuel generators with installed capacities up to 3.9 MW each, with similar maximum and minimum active power, power factor, ramp rates and fuel consumption in g/kWh. For this reason, the share of load served by each generator is quite similar, and any of these generators, which act as the base load generation, are used in the normal operation of the power system. It is important to highlight that even when flexibility analysis for stability control and other parameters need further analysis, some stability and operational conditions were considered. A stability study carried out in the power system of Isla de la Juventud (by the authors) consider certain operating conditions that should not be violated under any circumstances, although these results are sensitive and are not described in this article, they served as authors criteria constraint to reject certain results that did not meet these conditions. Thus, only the technically valid scenarios for the local conditions were prioritized in the analysis.

As a particular characteristic of the electric power system of Isla de la Juventud, residential sector is the largest consumer in the system due to the low activity in the service, industry and agriculture sectors. Consequently, today the peak demand is about 20 MW around 22:00 during a summer day, exceeding the demand for a winter day by 2 MW. In Isla de la Juventud, by the end of 2019, there was 15% of penetration by renewable energy into the electrical system, with a significant participation of PV (66.14% of RES) [27]. Until 2030, UNE has projected an increase with PV up to 15 MWp. Therefore, the difference between the load curve and the residual load curve is around 6 MW at maximum.

The analysis of the scenario RENES confirmed that the batteries would be an alternative to take advantage of the hours of maximum solar production by storing the energy to give the electrical power system a backup during the hours of maximum demand. Moreover, it is possible to increase the share of intermittent renewable power generation with the use of energy storages, such as electric batteries, as has been concluded in several studies of island isolated power system [33,34]. The results show the scope of RES introduction to the electric power system is technically possible and abundant solar, wind and availability to biomass can provide a transition up to 100% RES based systems, similarly to other islands in the Caribbean [35]. The MAXRES scenario supports the governmen<sup>t</sup> plan to increase the use of biomass for electricity production as Cuba has plans on promoting biofuel production nationally [7].

The growing concern about global climate change also drives the transition to nonfossil-based electricity production and the study shows that with increased RES the potential for climate mitigation is realized. The advantages of biomass utilization in the energy sector also support Cuba's sustainable development goals as stated by e.g., Bravo Hidalgo [36–38]. Similarly, decreasing costs, especially for wind and solar power, now offer a competitive alternative to conventional energy sources. Reducing the consumption of conventional sources towards zero decreases import dependence and supports sustainable energy transition necessary in Cuba to promote economic development [23].

The approach of the LINDA model enables the economic analysis of different scenarios in the view of future demand of the Isla de la Juventud power system and achieving different coverage with renewable sources. The LINDA model inputs include characteristics of the generation mix (also maximum and minimum required capacity of the fuel and diesel generators); sectoral economic development and sectoral energy intensity development; future growth rate for economy; future energy demand; installed and future power plant capacity and their respective load factors. The used Linda model version calculates the electricity consumption and production on an hourly basis and requires modifications if 15-min interval data is needed. The constructed model for Isla de la Juventud has only four economic sectors due to the data availability and can deal with structural changes within

the industrial sector only by modifying the whole sector intensity and growth figures. In this case the data availability is the problem because more detailed subsector data was not available. The model itself can include as many sub-sectors as needed.

For the scenarios constructed, however, subsequent flexibility studies are necessary e.g., to study if the transmission lines of the power system have the necessary capacity to transmit the energy from renewable sources and to avoid possible blackouts or curtailments. In addition, a stability analysis linked to system operation analysis is important to guarantee voltage and frequency stability, either in normal operating conditions or in the occurrence of a fault. The priority order of the power plants for producing the residual load is not using optimisation algorithms in LINDA model because information of the ramping rates and ramping costs of the different power plants was not available. The LINDA model does not include the transmission and distribution network and hence the distribution costs and potential bottlenecks are neither considered in the modelling.

The temporal and spatial resolution of the applied LINDA model ensures an adequate quality and validity of the results compared with other models. The temporal resolution of LINDA limits its scope to the evaluation of hourly demand and generation balancing by the sectors of the economy. The evaluation of the results indicates the study findings are quite accurate to real situation currently, although the data is incomplete and at times estimations of the parameters were used due to lack of accurate data. The sensitivity analysis or optimisation could not be carried out comprehensively due to availability or confidentiality of the data. Unlike other similar studies with LINDA, e.g., [22–24] the optimisation (especially least cost options) studies was found challenging due to complex trade agreements, heavily subsidized oil products and other factors that may distort sensitivity analysis and finding the "correct optimisation" results in Cuba.

This study with a LINDA model does not fully reflect operational restrictions, because it is based on a simplified technology representation to determine generation scheduling and reserve sizing. More detailed models need to be applied to reinforce the results of this work. The authors sugges<sup>t</sup> further analysis of e.g., Flextool to study the reliability, resilience and stability of the system, as the main challenges in integrating high shares of RES into the electrical system. Furthermore, the economic analysis of the lowest cost systems is an essential part of the analysis and equally important to the technological analysis of the modeling of the electric power system based on RES. This article focused on a set of technological possibilities, hence the authors recommend that further studies be carried out on the economics of the system on Isla de la Juventud. Only in this way will it be possible to identify the most suitable and lowest cost systems. An economic analysis of biofuel production was not carried out in this study, which provides another interesting area to examine the food-energy nexus: the costs and benefits of growing biomass for energy production versus food.
