Significance of Good O&M of Solar PV Power Plants

Good Operations and Maintenance (O&M) ensures high levels of technical, operational, safety, and economic performance of a photovoltaic system throughout its lifetime. These services are essential and mitigate potential risk, enhance generation and bring down the levelized cost of electricity (LCOE), thus, positively impacts the return on investment (ROI). These points are widely acknowledged by all the stakeholders in the solar industry.

Broadly speaking, the lifecycle of a PV project can be divided into 4 phases:

•        Development: few months to even few years
•        Construction: a few months 
•        O&M: typically 25+ years
•        Decommissioning, disposal and recycle: a few months

Clearly, the O&M phase is the longest one, therefore, increasing the quality of O&M services is crucial. The challenging part of O&M is not the Operations and Maintenance but is the ability to reflect the specialties of each system, topologies, installation sites, and country requirements. There is a lot of confusion or lack of clarity and knowledge in the industry regarding the minimum requirements (scope) which sometimes creates pain points down the line.

Some of the standards used during O&M activities are listed in the figure below:

Solar PV Power Plant’s Operation and Maintenance

Solar PV Power Plant’s operation used to be tractable in the past, but with increasing grid integration efforts, there is a requirement of more active & flexible operations by grid operators. Some activities like scheduled shutdowns, power curtailment, frequent adjustment of settings such as power factor, voltage and frequency tolerances demand customized operations to maintain the efficiency as well as safety of the plant. Thus, the operation process may look different for each plant but major operations involved in all the plants are:

• Maintenance of documentation and checklist of activities.
• Plant performance monitoring and supervision
• Performance analysis and improvement
• Reporting and Technical Asset Management
• Power plant security
• Optimization of O&M

The key to optimum operations is the efficiency of the management part in the plant throughout its lifetime. In a power plant, there are typically 3 types of maintenance:

1. Preventive maintenance: This is the core maintenance service which comprises regular visual and physical inspections, as well as verification of all the activities in the plant. This maintenance is carried out in regular intervals decided by the OEM and O&M planning committee and are included in the annual maintenance plan.

An example of preventive maintenance is thermographic inspection which aims to identify defective panels on a PV plant. Several defects like hotspots, moisture ingress, soiling, etc can occur on the module, significantly reducing the productivity of the modules. Therefore, relevant inspection procedures are performed either by hand-held cameras or by using drone technology.

2. Corrective maintenance:

It corresponds to any activity performed to restore a PV plant system, equipment, or component to an earlier functioning state. This occurs after fault detection by remote monitoring or during an on-site inspection. These activities includes:

• Fault diagnosis - to identify cause of fault and locating it.
• Temporary repair - to restore the rquired function of a faulty item for a limited time, until a permanent repair is carried out.
• Permanent repair - to resolve the fault permanently.

3. Predictive maintenance:

This is a condition-based maintenance carried out due to the results from analysis and evaluation of the significant parameters of the degradation of an item. The operations team does predictive maintenance through regular monitoring, supervision, forecast, and performance data analysis.

Data Analysis for Predictive maintenance

Since predictive analysis is forecasting derived activity, there is a constant need for data analysis to ensure the well-being of the plant. The different types of data analysis done on the plant are:

• Breakdown analysis: This analysis can help identify any specific problem of different components. Breakdowns can be divided in 2 types - Forced breakdown and Scheduled breakdown. Forced breakdown/unwanted breakdown can occur because of poor device quality, poor installation quality, inadequate maintenance, loose termination, voltage surges due to lightning, etc., while scheduled breakdown occurs during any planned maintenance.
• Historical Generation and Performance Analysis: This analysis helps in forecasting the future performance of the power plant. This includes monthly, quarterly, annual as well as year-on-year analysis. It also included analysis of different losses in the plant e.g. AC losses, Auxiliary losses, Plant unavailability, Inverter efficiency losses, etc, which also detailed further in Loss Analysis
• Inverter Efficiency Analysis: Inverter efficiency is a very important parameter which is often not monitored. It is simply the ratio of Inverter’s output by input, but there are two ways to calculate it: power efficiency and energy efficiency. Power efficiency is the instantaneous efficiency of the inverter and various parameters like %loading, operating temperature, etc. The efficiency profile (%efficiency vs %loading) can be drawn by calculating inverter efficiency at multiple time-intervals and any abnormality in this curve can directly indicate the inverter’s performance. Similarly energy efficiency tells the average efficiency of the inverter for a certain time period..
• Loss Analysis: This includes the analysis of many different types of losses like shading loss, IAM loss, soiling loss, temperature loss, mismatch loss, module degradation loss, DC-AC ohmic loss, auxiliary consumption, transformer loss, system unavailability loss and transmission line loss.
• SCB and string analysis: SCB and string data analysis is also often neglected. Either the solar plant does not have string/SCB level monitoring or even in case monitoring is present, there are high chances that no analytics is done on this data. Not only string/SCB level monitoring is helpful in indicating any string/ SCB faults in real time for quick debugging, but also if this data is analyzed after proper normalization, it can even predict the underperforming strings/SCBs..
• WMS analysis: WMS is one of the most crucial instruments installed in solar plants. A weather monitoring station can be immensely helpful in monitoring the performance of any solar power plant. This instrument is installed to monitor important parameters like GHI & POA (Plane of Array) irradiation, ambient & module temperature, relative humidity, wind speed & direction, rain accumulation, etc.

O&M activities often multifolds when unpredictable events like natural disasters, theft, fire, equipment damage, series of failures, deterioration due to design faults, or even new design modifications required due to regulatory changes, etc. occur. Other additional maintenance responsibilities include module cleaning, vegetation control, general site maintenance, waste disposal, drainage, building maintenance, on-site measurements, etc.

Even though there are so many steps taken to maintain the performance of the plant, we have observed many issues on-site. The reasons may be ignorance, insufficient information availability, or miscommunication on-site. Whatever the reason, it directly hampers the performance of the plant. We are sharing one such case study below:

Wrong connection of PV array in the SCB

In a plant of 50 MW, we recently observed that there were wrong connections of PV arrays in the SCBs which made the maintenance of specific PV arrays a tedious task.

Brief background:

In a SCB, strings are terminated usually via a connector (shown by numbers in the below figure) which contains a label of the string connected. So both positive & negative termination of the PV array / string is to be done as per correct label numbers (Figure 1).

But in this plant, these connections are not done as per correct label number. Rather, the positive and negative terminals from the same array were connected to different label numbers in the SCB as shown in figure 2. In this scenario, one could argue that the connection is ok because all the arrays are connected in parallel.

But what could go wrong? Imagine you need to disconnect array 1 from the system (in figure 2). If the O&M team will remove the connector no. 1 from both positive & negative sides of SCB, they will be under the impression that the array 1 has been disconnected. But this will not be true, since the connections were not right in the first place. 

The above figure doesn’t justify the complexity of the issue. Imagine there are 16 or 32 strings in the SCB and all the connections are mixed-up. It will take you an ample amount of time to debug these connections. And imagine doing that in adverse times like during array faults, which can cause delay and can also lead to SCB damage.

Our team observed that this problem existed in almost all of the SCBs in a 50 MWp plant. Which compels us to think of the frustration of the O&M team in times of fault rectification in the plant!

- Pranav Maheshwari, Co-Founder and Head of Technology, PV Diagnostics

- Sudarshan Bhosle Lead Analyst - Backend Analysis team, PV Diagnostics