System for performance assessment of solar home systems

Research article. https://doi.org/10.16925/2357-6014.2021.02.01 1 Universidad Autónoma de Bucaramanga, Colombia. Email: ymunoz294@unab.edu.co ORCID: https://orcid.org/0000-0002-5213-1884 https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh= 0001478388) 2 Universidad Autónoma de Bucaramanga, Colombia. Email: mpinto12@unab.edu.co ORCID: https://orcid.org/0000-0002-2181-5546 https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh= 0000156056) 3 Universidad Autónoma de Bucaramanga, Colombia. Email: cvera678@unab.edu.co ORCID: https://orcid.org/0000-0001-8593-9567 https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh= 0001647505) System for performance assessment of solar home systems


INTRODUCTION
In recent years, renewable energy projects such as solar, wind, hydro and biomass have begun attracting worldwide attention and replacing energy generation from fossil fuels such as oil, coal and gas [1], leading to its implementation being widely encouraged by their integration advantages as a solution to the increase in energy demand and a need to reduce greenhouse gas emissions [2]. Solar photovoltaic (PV) energy is considered a good alternative for rural remote areas where there is no power grid and for off-grid applications at a small scale; socalled solar home systems (SHS), which are used to supply the lack of electric power in developing countries [3]. A typical SHS consists of a solar generator (PV module), inverter, battery bank and charge controller, as well as the connected appliances. Solar PV modules charge the batteries during the daytime to supply the night consumption, while the charge controller manages the input and output energy of the battery bank [4], [5].
Colombia is a developing country that receives solar radiation almost throughout the year with an annual mean of 4.5 kWh/m 2 /day over its territory [6]- [8]. Despite having a huge scope of solar energy generation, there is still a lack of energy supply with about 51% of the national territory being formed of remote rural or non-interconnected zones (ZNI for its abbreviation in Spanish) [9]. Less than 3% of the population in the country is supplied with this type of energy, thereby identifying a deficit in the three elements that drive PV development in a country: politics, research and monitoring [10].
This article presents the design of a PV test system for off-grid solar home system (SHS) performance assessment, analyzing a case study based on the behavior of a user's daily demand in the rural area Hato Corozal, Casanare.

Literature review or research background
Considering that photovoltaic systems are characterized by their easy installation, modular installations and energy independence provided to users [11], [12] they represent a viable alternative for grid-tied and off-grid PV systems that were already in operation and located in remote areas [13]- [15]. Shiva & Sudhakar (2015) presented a performance analysis of 10 MW photovoltaic system connected to a power grid and analyzed the performance of a large grid-tied system in operation as a parameter that could assist in the design, operation and maintenance of new systems connected to the electrical grid, validating experimental results with the support of software such as PVSyst and SolarGis [13]. Querikiol  using Homer Pro software for optimization of the solar system [14]. Aghenta & Iqbal (2019) present the design of the PV system analyzing in detail the user's loads, number of appliances, number of people, location, solar resource, weather variables and the integration of other energy sources such as diesel generation [15].
An existing solar PV 200 kW power plant in India was studied for its off-grid and on-grid configurations in order to analyze their performance parameters through measured data such as annual energy yield [16]. Muhammad et al. (2018) simulated in PVsyst and compared the performance that a stand-alone PV system would have in two geographical sites (Quetta, Pakistan and Copiapó-Chamonate, Chile), analyzing performance ratio, solar fraction, energy supplied to the load, etc. [2].
There are also several projects developed in rural or non-interconnected zones [10], [17], [18], however, it is not common to find a performance assessment of the system in its different stages, while in literature research there is not enough information available about the performance of solar home systems [19]. In this way it is difficult to know if the installed systems are functioning efficiently, thus it is being assumed that sizing criteria are precise.
It is also important to consider that effective component sizing ensures a reliable, proper and economical design [20]- [23]. Monitoring the performance and loss factors of PV systems is important to assess overall efficiency and to enhance productivity [24], thus a successful implementation of solar PV systems involves knowledge on their operational performance under varying climatic condition [25].

Energy requirement assessment
To determine the daily energy requirement, E daily (in Wh/day), it was necessary to identify all the household electrical loads with their wattage and the mean amount of time they were used in a day as shown in Table 1. An approximate total power consumption of 0.90 kW and a daily energy requirement of 2.54 kWh/day was obtained, which coincides with the value monitored in Hato Corozal, Colombia [26] and the 2.67 kWh/day reported for rural users (strata 1) in Casanare department [27].

Load monitoring
Once the sizing and selection of the PV system components were carried out, loads were programmed by timers that allowed for them to be turned on/off at specific times to emulate the daily load curve. Figure 1a was followed for the connection of the components with the PentaMetric monitoring tool to analyze system performance, which has a three shunt that measures hourly current data and two connectors that measure voltage, energy and temperature from the batteries and PV modules. For the inverter, only current was measured, so the energy was calculated considering that the inverter voltage is the same as the battery bank voltage. Solar radiation was also monitored with a weather station available in the renewable energy laboratory of the Universidad Autónoma de Bucaramanga, UNAB and monthly data were obtained.
Two cases were analyzed, for the first one or Case-1 the total daily consumption of a user was considered without taking into account the behavior of their electrical loads and the hours when their consumption peaks are presented, while in Case-2 the user shifts their highest consumption to the daytime, seeking to coincide them with the hours of higher solar radiation so that it adapts to the characteristic load curve of the area.

Materials
The inverter charger used is a TECA IIP-241000BF with 1 kW of rated power and 3 kW of surge power, which allows for connection to the power grid to receive energy and to charge the batteries while supplying the demand of the charges. voltage to 24 V and two in parallel to increase the capacity of the battery bank [28].
The solar charge controller is the Acacia ICM-4024150 with a maximum current capacity of 40 A. The final installation and specifications of the PV system are shown in Figure 1b and Table 2, respectively. The performance ratio PR considers the possible losses of the system and it was also calculated [29], [30].

Figure 1. PV system layout (a) and final installation (b)
Source: The Author.

RESULTS
According to the behavior of the proposed system, the mean energy consumption was 2.41 kWh/day. For Case-1, Figure 2   SoC was 79%. This means the DOD was not deeper than that specified in the sizing.
The peak voltage and highest SoC was presented at 8:30 h with values of 28.4 V and 118% respectively, which happens one hour after the highest generation, as shown in Figure 2.
For Case-2, Figure 3 shows a peak generation of 738 Wh, reached at 10:00 h with the highest solar irradiance of 742 W/m 2 and a PV performance ratio of 0.95. The hours of highest solar irradiance (from 9:00 am -12:00 pm) presented a high performance factor, due to the coinciding consumption peaks. The PV supply and radiation relationship were favorable for the system due to the solar resource was used in a greater proportion; it means the PR was better than in Case-1.

DISCUSSION AND CONCLUSIONS
A system for performance assessment of solar home systems (SHS) was designed, sized and implemented for monitoring their performance parameters through measured data such as solar radiation, a battery's state of charge (SoC) and PV generation since most of the studies have been done either for grid-tied or off-grid PV systems already in operation.
The sized system satisfies the energy requirement without a DOD deeper than expected. In Case-1 a PR of 0.77 was obtained, nevertheless this value decreased to 0.70 despite having a good solar resource due to a low energy consumption by the loads and the full battery bank SoC; this then increased when the energy required exceed the PV generation. In Case-2 the PR was 0.95. In both cases, the SoC was higher than 79%.
It was found that the excess solar energy not demanded or stored by the system affects its performance in a negative way due to an undersized battery bank, therefore it is necessary to expand the capacity of the battery bank.   A higher performance of the system is obtained when considering an hourly consumption and load shifting during its operation in such a way that the consumption peaks coincide with the hours of higher solar radiation, as this takes advantage of its generation potential (ratio of available radiation and energy generated). However, in this study the SoC of the batteries did not vary considerably between an hourly distribution of charges and a total daily consumption value because the nocturnal consumption peaks were not greater than the daytime ones; for this reason, in the event of a load curve with peaks centered on night hours, the proposed charge values could not be guaranteed and the useful life of the batteries could be affected.
This system is configurable to operate with the electrical network in case it extends to the rural zone, offering interactivity with the network to develop smart grid strategies. Findings in this study suggest that sizing with appropriate performance parameters helps in optimizing factors such as associated costs, component operation and lifespan.