The majority of the population in Sub-Saharan Africa are without access to clean electricity amongst other basic human amenities, with most of the population located far from the conventional power grid. This means that poverty is at its highest in these developing countries due to lack of power, which could be the key driver to economic transformation. Decentralized generation sources through microgrids have emerged as a means of generating and utilizing power locally. Efforts to deliver power through SPV microgrids has met a rather unanticipated halt due to comparably low lifetime in the field, bringing into question their reliability and effectiveness. The main problems identified were classified as: design, planning and management issues, O&M, market constraints, lack of skills & knowledge, and policy frameworks. This is supported by a number of studies (Case and Ketlogetswe 2016; Akinyele, Belikov, and Levron 2018; Njoh et al., 2019). It is important to address these to benefit from the many advantages brought by microgrids.

The continuous population growth, as well as expansion in industrial activities, has created an exorbitant energy demand. Conventional power generation methods have also become ill-fated due to the issues of global warming and environmental pollution they render and, coupled with the depletion they are facing, are non-renewable in nature (Machrafi 2012). Renewable energy applications are gaining support and recognition in the quest for decarbonization of the energy sector and reducing carbon emissions since they are unlimited and naturally replenished (Messenger and Yogi Goswami, 2015; Kale 2016). SPVs are one of the fastest growing forms of renewable energy generation in terms of development and technology advancement. According to a report on “Future of Photovoltaics” by International Renewable Energy Agency (IRENA), the SPV installation capacity has increased from 40 GW to a massive 583.5 GW within the span of 10 years (International Renewable Energy Agency—IRENA 2019). This is consistent with the work of Gentries (Bradford 2006) on the evolution of solar PV technology dissemination. The author argues that, due to the recent intensification of research and development on SPV technologies to yield cost-effective and feasible energy solutions, SPV applications are tipped to take over domestic and commercial markets. This, however, is reliant on reliability and optimum system performance, specifically energy yield, which is pivotal to ensuring the success and continual deployment of SPV technology, whether for small-scale or for utility grid applications (Jahn 2003).

In its nature, SPV power generation is surrounded by uncertainty due to intermittency issues, hence giving rise to specific operational and planning problems. This uncertainty means that when forecasting energy yield, accurate data is needed to produce reliable and dependable estimates (Ventura and Tina 2016). While solar microgrids may intrinsically be low-risk systems as compared to the conventional power generation setups, they are subject to unique risks associated with financial, operational, and technical aspects (Baurzhan and Jenkins 2016). Regular problems in SPV systems emerge as reduction in efficiency and output deficit stemming from faulty events. These therefore prompt diligent and thorough inquiries on possible fault types, failure mechanisms, as well as detection and classification to perform remedial actions. Research is abound with studies pertaining to fault analysis in SPV systems as a necessary evil to protect system components from catastrophic failure (Zhao et al., 2013; Haque et al., 2019) with various fault localization and detection techniques developed (Chine et al., 2014; Silvestre et al., 2014; Zhao et al., 2015; Dhimish et al., 2017; Harrou et al., 2019).

Tragic experiences emanating from faulty conditions in SPV systems have been studied in the past. In previous research (Eltawil and Zhao 2010; Kandpal et al., 2016) an assessment of PV module failure was carried out, identifying some of the common fault scenarios including potential induced degradation, cell cracks, and encapsulant discoloration. The understanding of faults and failure modes plays a critical role in deciding the best method to employ in dealing with the situation to prevent further or impending catastrophe (Osmani et al., 2020). To deal with this, sensors, control, and protection devices are hence installed and incorporated in the system for monitoring purposes as well as facilitating the smooth functioning of the network (Louie 2018). Various research has been carried out to address these various topics relating to the general approach to SPV microgrids’ operation at a high level and careful scrutiny with emphasis on areas of optimization of system design and energy production forecasting, system performance, reliability, and safety (Tahri et al., 2013; Allamehzadeh 2017; Dhillon 2017; Oprea et al., 2019) as well as operation and maintenance (C 2010; Paul and Bray 2012; Woyte and Goy 2017; United States Agency for International Development 2013).

The incentives and policies in various countries have also put more weight on the need for performance assurance, which is warranted by adaptation of standardized and pertinent O&M approaches in SPV systems. Moreover, this standardized approach facilitates planning and decision making as well as much needed guarantee of sustainability and long-term impacts by key players or stakeholders in the SPV microgrid industry (Lynn 2005). The approach comprises of a maintenance strategy with clear techniques, measures, and procedures to ensure maximum system availability and ongoing optimal power generation. Maintenance strategies tend to be industry specific i.e., being better suited for particular kinds of equipment or machinery in that sector. The SPV industry encounters particular fault events, as has been highlighted. Therefore, a specialized maintenance strategy needs to be adopted to ensure stable and safe operation over the system lifespan of 20 years and beyond.

The significance of O&M of SPV microgrids is discussed next, followed by a brief overview of the operation of solar photovoltaic microgrids in the next section, giving an idea of the general layout of the system and the main aspects of operations, i.e., planning maintenance strategies, supervision, and control as well as plant performance monitoring. This is followed by further discussions on the various maintenance strategies employed in the field: corrective maintenance, preventive maintenance, condition-based maintenance, and predictive maintenance. Subsequently, attention is drawn to some key features and merits as well as limitations of each stratagem to stress out the serious decisions required in selection of which approach to adopt. More work is reviewed on the elements of a successful O&M program and the way to go about developing it. The review then proceeds to look at the three models of executing an operation and maintenance program: In-house O&M team, Third Party Contract (O&M service company), or Installation company (EPC company). Finally, a conclusion is made, and future works are put into perspective. Figure 1 shows the structure of the manuscript.

The report by C. Marney & Co. entitled “Microgrid evolution roadmap” defines a microgrid as “electricity distribution systems containing loads and distributed energy resources, that can be operated in a controlled, coordinated way either while connected to the main power network or while islanded” (Marnay et al., 2015). This definition portrays the characteristic features of the microgrid as described by Schnitzer et al. (2014), i.e. micro-grids are capable of generating power locally and supplying electricity to a relatively small number of users who are connected to each other through a shared distribution system. Essentially these serve as the main distinction between a standard SPV system, such as rooftop solar for an individual household, which is comprised of the basic SPV components. Furthermore, the high investment cost for a microgrid which includes the transmission through complex network cables, power modulation, and other power electronics for enhanced control and protection renders an extra requirement for operational and maintenance needs. The diverse microgrid architectures utilized in practice coupled with these seemingly passive components presents control issues associated with increased inspection and maintenance activities.

The SPVM operates as an isolated and autonomous power distribution network mostly for rural and remote community settings relying on solar as the main power source (Schnitzer et al., 2014). The system may be coupled with another source (usually diesel) as a backup or even battery power storage system (capital intensive) for uninterrupted power supply to the end-user (Katiraei et al., 2007). SPVM offer flexibility in the sense that it can take different configurational structures where there can be a central generation point or various points of generation in a household which serve as suppliers in different sectors of the area but combined into a single transmission network for demand management (Buchholz and Schwaegerl 2004; Hirsch et al., 2018). Figure 2 is a graphical representation of the main elemental aspects of a microgrid network based on functional classification. Portrayed by the block diagram is the characteristic features of a microgrid in general distribution systems, showing a microgrid as distinct from the rest of the system. It also displays s the power production resources and its ability to function exclusively without connectivity to a larger grid (Hirsch et al., 2018; Global Sustainable Energy Solutions GSES (2017)).

Nicholas. O, 2019, carried out a comparative study of DC-versus AC-based microgrids. The study highlights the modularity of microgeneration systems as a great characteristic advantage in rural setup(microgrids) as it supports ease of modification in response to varying energy demands (Opiyo 2019a). SPV systems located on different houses can be integrated together through a web-like network microgrid. While this setup brings more reliability and robustness, this also comes with protection and control challenges associated with increased interconnections (Zhao et al., 2013). PV systems are like any other electrical power generating systems, as such, it is critical to follow the appropriate electrical standards, codes and principles of operation and interfacing when developing these microgrid networks for safety and reliability (Argetsinger and Inskeep 2017; Forum, Economic, 2020).

A solar PV system as depicted in Figure 3 is a synergy of individual components which includes solar panels, charge controller, battery (power storage), charge controller, inverter and power transmission cables, as well as electrical loads (appliances) (Ali et al., 2019). There are various ways in which to connect the components to supply the consumers/loads; this is referred to as network topology (Sebitosi et al., 2006). Illustrated in Figure 4 are the variable ways of connection that reflect the flexibility of setting up the network depending on the geography of the settlement, need to separate out loads of different priority, the resources, or the anticipated operation model (Bhattacharyya and Palit, 2014).

According to Goswami, the lack of motorized equipment in SPV systems renders them anti-operation and maintenance free, but that is not actually the case (Goswami and Kreider 2001; Messenger and Yogi Goswami, 2015). SPV systems are subject to various faults leading to power losses and consequently lack of returns on investment. Essentially, solar power generation needs to be tapped on to the highest order i.e., maximum energy yield and efficient performance when solar is available because of its dependence on weather conditions (Paul and Bray 2012; Allamehzadeh 2017) The smooth flow and operation of SPV systems is made possible through the use of control and protection devices (Park et al., 2009; Yang et al.,2014; Salman et al., 2018). Their function also encompasses detection and isolation of faulty components from the system, thereby minimizing energy losses and improving the power supply robustness and system reliability in general (Gokmen et al., 2012; Alam et al., 2015). Data acquisition is a fundamental aspect in tracking optimum operation, energy generation forecasting, and optimization of system performance MPPT. This could be real-time system data or historical performance data utilized in achievement of various objectives and making informed decision (Drews et al., 2007; C 2010). In radial configuration, the DC bus supplies power to the loads through a single path, usually where there are low voltage demands to avoid extra conversion stages. There is also the ring configuration, also termed loop network, whereby there are multiple paths between the source (DC bus) and the consumer, whereas in a tree configuration the paths of power from the DC bus forms branch-like lines before further spreading to different loads connected to them (Kumar et al., 2017; Asian Development Bank 2020).

According to (British Standards Institution BSI, (2010))—BS EN 13306:2010, system operation is referred to in clause 2.9 as “The combination of all technical and administrative actions intended to enable an item to perform a required function”, recognizing necessary adaptation to changes in external conditions. Operations of solar PV microgrids encompass some key processes which complement or work together for the optimal system upkeep, reliable power supply, and improved system efficiency while achieving system longevity and reduction in maintenance costs. These are monitoring (remote or onsite), supervision and control, and planning and coordination of maintenance strategies (Practices and Version, n.d.). A summary of the diverse viewpoints to SPV plant operations is portrayed in Figure 5.

SPV power plants are, in most cases, unmanned and remotely supervised; therefore, remote monitoring is crucial for enhanced performance as well as reliability benefits (Tejwani et al., 2014). Operational monitoring is an essential part of the service life of SPV plants from system commissioning until the disposal stage or decommissioning phase. At this juncture performance monitoring serves to evaluate and analyze the working conditions of the solar PV installation to unearth and track down any design problems or anomalies in the system (Belhadj-Yahya 2010; Paul and Bray 2012). It can be said then that operation monitoring is a necessary ingredient for quality and safety assurance. Not only that, it is paramount in the optimization of output power since solar is susceptible to fluctuations due to weather and other unavoidable operational losses(Belhadj-Yahya 2010).

By discussing the various microgrid modes or SPV system architectures and configurations, the implications—especially on the failure modes and ultimately the maintenance needs - are consequently unearthed for clarity in dealing with such.

This maintenance strategy is also known as reactive maintenance; it entails unscheduled remedial actions undertaken to rectify failures, breakdowns, or any signs of inability to perform a function so as to restore a system or equipment to optimal performance (Dhillon 2017; Solar Power Europe [SPE] 2018).In a reactive mode, the promptness coming with being required to repair failure conditions results in diminutive attention paid to ensuring that the operating conditions are within the allowable specifications, resulting in the actual service performance and life-span of the equipment being compromised. It is a risky type of maintenance rendered by the fact that pressure may be on the maintenance team to return the system to normal operation quickly, giving them less planning time. Even though not planned for, in some instances corrective maintenance activities can be scheduled for during night or low irradiation hours to avoid interruption in system operation (Myeni 2019; Woyte and Goy 2017). During repair works we normally consider whether to repair or replace. This decision relies on availability of replacement parts, lead time for replacement parts, cost of replacement parts, and difficulty and/or complexity of repair requirements. The primary goal of CM is to repair and restore equipment to its normal condition after failure has occurred.

Preventative Maintenance entails the precautionary steps undertaken for forestalling or lowering the probability of failure or an unacceptable level of degradation (Dhillon 2002).Scheduled maintenance is characterized by strategically planned, intervallic, and specific scheduled intervention measures to maintain equipment in the specified operating condition through the process of checking and reconditioning. Routine upkeep actions are carried out according to and with careful study of manufacturer’s recommendation regardless of the actual performance and condition of the asset (Van Horenbeek et al., 2013). A critical aspect that needs careful attention with this strategy is the ideal length and time for maintenance intervals and the actions that may be performed at variable, systematic, or based on the condition of the system(Alam et al., 2015; Baklouti et al., 2020). Systematic preventive maintenance is characterized by replacing aging or unfit components according to a predetermined interval. The condition-based maintenance approach is carried out subject to the state of the components i.e., when signs of incipient failure or abnormal performance are observed. Even issues of performance from aging can prompt components’ replacement (Sharma and Chandel 2013). In their study Said and Walwil (2014), undertook a study on the dust fouling factors on PV modules performance, showing a reduction in spectral transmittance at 30% representing irradiance to the cells and hence less power conversion. As proactive measures, the study proposed, as a means of mitigating soiling, mounting modules at an angle and anti-reflective coating on glass covers. The status data based on equipment performance monitoring and the control is analyzed and corrective actions proposed. PM is meant to curtail system performance decline and obsolescence risks by minimizing the chances of failure occurrence as well as accelerated components’ degradation.

With this maintenance method, appropriate mitigation measures are taken (repair or replacement) following the prediction of possible system failure. Remedial actions are performed to forecast and are recommended when maintenance procedures necessary to deal with any incipient failure or fault condition are observed (Li et al.,2016). This observation is centered on thorough, continuous, and regular monitoring, supervision, forecast, and performance data on historical performance and anomalies for analysis of the SPV plant (Bosman et al., 2020). The key is in identifying hidden anomalies and irregularities before the next circuit testing or thermal imaging inspection which indicate upcoming component or system failures or underperformance (e.g., at PV modules, inverters, combiner boxes, trackers etc.) (Chebel-Morello et al., 2012; Outline 2015). PdM is a proactive approach that deals with faults and failure conditions which target and eliminate symptoms of failure—carried out by integrating analysis, measurement, and periodic test activities.

Choosing which maintenance strategy to adopt for your SPV system is not an easy decision to make; it requires an all-round understanding of the various contributing factors and the special maintenance needs of the system at hand i.e., SPV system.

As shown in Table 1, the choice of maintenance approach is a tough one requiring some balances and tradeoffs to meet the set goals and objectives of the maintenance team. Some key goals include longer system uptime or availability and maximum energy yield. Essentially the first integral and determining factors comes in the form of available resources which are: financial budgets, labor, knowledge and skillset of operators, and extent of energy requirement (i.e., is the system serving as backup or main source of power). A good maintenance approach is characterized by a mix and match of the various strategies discussed to get the best benefits of the traits outlined and it is established based on risks and cost-benefit principles, substantiated by literature on Table 2. The decision here is of critical importance before developing the actual O&M plan so that tasks of scheduling appropriate skill requirement, predicting future spare part consumption, unequivocal in the maintenance actions, and any specialist work that may need to be outsourced are prepared for.

It is common knowledge that how renewables perform in general is influenced by the environment within which they are deployed. This is the case for SPV systems, and several studies have in the past been undertaken to analyze this key subject area (Awwad et al., 2013; Bouraiou et al., 2015; Bonkaney et al., 2017; Abdalla et al., 2019). Extensive research on the performance of PV modules installed in the desert region in South of Algeria has also been performed. Taking this into consideration, the operational needs of the particular SPV system are influenced immensely by the type of climate within which it is installed (Mustafa et al., 2020). An O&M plan must be suited to the environment condition with which the SPV plant in question is installed; this is to be as optimized and cost effective as possible, for instance anti-defrost measures may not be included for SPV systems on hot climates throughout the year.

Growth in the deployment of solar PV technologies in aggregate has indirectly resulted in advancement in maintenance practices. Researchers have developed cutting edge technologies and tools for improved and accurate system fault diagnostics, troubleshooting, and prompt remedial actions response by integrating options for document management, technical reporting, fault analysis, maintenance scheduling, and financial reporting (Gilabert and Arnaiz 2006; Efthymiou et al., 2012; Abbassi et al., 2015). O&M being a heavily technical aspect in solar PV technology is reliant on the size of a system, specific design, and location context. These technicalities cause a variation in maintenance implementation approaches from place to place and the different installation requirements. A maintenance management framework hence is necessary to ensure remedial actions are employed as and when needed to avoid downtime and reduction in service lifetime of the overall system. The key areas with respect to developing an all-round maintenance program are highlighted in Figure 6.

Mgonja, C. T. and Saidi H. undertook a study focusing on off-grid SPV systems in public facilities (Mgonja and Saidi 2017), with a keen eye on maintenance management systems to evaluate effectiveness on its implementation. This evaluation was based on a case study method to propose a model that reflects the need for collective efforts by all relevant stakeholders involved in SPV systems from planning to disposal phase. The proposed model is aimed at improving field maintenance practices by enforcement of active participation by stakeholders in fostering proactive maintenance culture and highlights that failure to swiftly implement a maintenance management scheme in the early project phase has a direct impact on long-term sustainability. In British Standards Institution (BSI, 2010), it is reiterated that, from project conception through to construction, supervision, and inspections as well as commissioning tasks, continual and extended planning is vital for the success of solar PV installation.

Responsibility fragmentation has a direct contribution to the success of O&M of SPV microgrids. This is possible only if the workflow, from early identification of incipient failure to prompt response by deploying maintenance personnel with the right tools and spare parts, is a combined effort of the asset managers, plant operators, and in-field technicians (Averbukh et al., 2013; Tejwani et al., 2014). As is seen from the Schematic for the main aspects of a maintenance program, the operator is solely involved with O&M aspects, leaving the microgrid managerial and finance side of operations to the Asset Manager (Palit and Sarangi 2014).As a result, there is assurance of sustainability because of the accountability instilled by this model (approach).