Skip to content

Operation & Maintenance Best Practices Guidelines (Version 6.0)

Do you prefer the guidelines as a pdf file?

Download PDF

Are you interested in downloading a specific chapter?

13

O&M for rooftop solar

This chapter is to assist in the application of established utility-scale best practices, detailed in the previous chapters of the document, to rooftop solar projects. It also highlights where rooftop solar projects are distinctively different from utility-scale projects, and where they may require specific O&M best practices that may not be present or applicable for utility-scale projects.

A rooftop solar PV system has its electricity-generating solar panels mounted on the rooftop of a residential or commercial building or structure. On residential buildings they have typically a power of about 5 to 20 kWp, while those mounted on commercial buildings often reach 100 kWp to 1 MWp. Large rooves can house industrial scale solar PV systems in the range of 1-10 MWp. Since O&M organisation depends on size and structure of the asset we distinguish between:

  • C&I (commercial and industrial) rooftop solar and
  • Huge portfolios of residential systems (distributed solar portfolios)

13.1. C&I rooftop solar

C&I rooftop solar systems are designed and installed for commercial or industrial applications. They are either built, owned, and operated by an IPP who then sells electricity to a company or institution via a PPA, or ownership is transferred to a company or institution by an IPP which continues to operate the installation. In addition, a growing trend observed internationally in energy-intensive built-up urban areas with high-rise residential, commercial, or mixed-use blocks, is for rooftop solar systems that either feed into the grid or are set up as distributed generation.

C&I rooftop solar systems frequently occur in what is known as a “distributed generation” setup. Distributed generation refers to energy-generating technologies, including solar solar PV, that are sited either on or nearby the premises that are consuming that energy generated. Sometimes distributed generation energy systems are part of a microgrid that offers a degree of crucial power independence from the main grid in cases such as mains electricity outages during extreme climate events. C&I distributed generation is being paired increasingly with on-site energy storage solutions to enhance energy independence and efficiency for the site.

Due to the relatively significant size of C&I rooftop systems (500kWp-10 MWp), the best practices highlighted elsewhere in these Guidelines should be applied to these installations. However, their location on rooves and their situation in commercial/industrial environments require additional guidelines to address these factors.

Regarding H&S considerations for C&I rooftop solar, the necessary precautions outlined in Chapter 2. Health, Safety, Security, and Environment should be taken into account, but need to be complemented to address the dangers associated with working at height (see for example Best practice guidelines for working at height in New Zealand, HSA Guide to the Safety, Health and Welfare at Work or IACS Guidelines for Working at Height). These additional precautions include:

  • Presence of permanent guardrails or other forms of edge protection
  • Presence of maintenance corridors
  • Use of mobile elevating work platforms, forklift platforms, etc.
  • Use of safety mesh
  • Use of temporary work platforms (also to avoid damage of modules)
  • Marking of dangerous areas (for example, fragile roof material)
  • Correct use of harness systems and lifelines
  • Correct use of ladders

As a best practice, aerial inspections should be conducted with drones as they can safely and accurately capture visual and thermographic data in significantly less time than it takes to manually inspect the entire array. The images and reports generated allow O&M technicians to identify precise locations and even the types of rectifications required. This enables swifter, safer repairs that also save costs in terms of repair time required.

13.1.1. Operations

An asset-centric approach to operations that promotes the free flow of data and transparency between all stakeholders for the entire lifecycle of the asset should be followed. This is made possible by using a monitoring and asset management platform.

Operating a C&I rooftop solar asset is similar in principle to the guidelines mentioned in Chapter 5. Power Plant Operation. To recap, it should include:

  • A Document Management system
  • Plant performance monitoring and supervision
  • Performance analysis and improvement
  • Optimization of O&M
  • Maintenance scheduling
  • Spare part Management
  • Decommissioning

As best practice, drones equipped with visual, thermographic, and other specialised inspection equipment should be used to support the O&M operations of C&I rooftop solar assets. They can provide image data that can identify anomalies missed by ground monitoring equipment. This allows problems to be spotted and rectified in a proactive, time- and cost-efficient manner, reducing the likelihood of more serious issues further down the line.

To accurately calculate the Energy Performance Index, collection of Reference Yield (Local Irradiation) and temperature data is required.

The following methods can be applied for collection of reference yield:

REFERENCE YIELD SOURCEACCURACYHARDWARE COSTCOMMENT
Onsite PyranometerHigh*HighFor more information, see section 9.10.1 Irradiance measurments. Public pyranometers may be used if available
Module level sensorHighHigh
Satellite DataMedium-High**None  For more information, see section 9.10.1 Irradiance measurments.
Cell SensorsMediumLow
Local ComparisonMedium-LowNoneThe established baseline must be verified
Historic DataLowNoneMonthly variation must be +/- 20%

*Pyranometers and cell sensors need periodical cleaning and recalibration to keep the highest level of accuracy. If this cannot be sustained, a good satellite irradiation data set is preferable.

** Satellite data accuracy depends on type of source. However, the best references have a granularity of 3x3 km² and do not include local shades. It is also worth noting that real-time satellite data provision comes at a cost. Another alternative is comparing the performance of neighbouring systems

TABLE 12 - METHODS SUGGESTED FOR THE COLLECTION OF REFERENCE YIELD

The variety of conditions leads to a higher incidence of uncertainty: greater shade, lower data accuracy, lower comparability between assets.

For example, greater and more variable shade profiles, due to significant roof obstacles, require that expected yields used in the EPI are adjusted based on shade expectation for the KPI interval.

As shading and vegetation control tend to be an ongoing problem for smaller-scale C&I given their relative size, and proximity to trees and gardens, as well as ongoing construction of neighbouring buildings that could affect the shading profile of the solar PV installation site, drones can be considered as a fast, accurate, safe and non-intrusive method of delivering shading analysis and vegetation management inspections at regular intervals appropriate to the site.

As a recommendation, horizon and obstacle plotting should be included in all yield modelling.

13.1.2. Maintenance

C&I O&M service providers should provide a Maintenance Plan to the Asset Owner during or before system commissioning.

Roofs under warranty require annual preventive roof maintenance to maintain the roof warranty. It is best practice for the retailer/installer and O&M service provider to meet with the roof maintenance provider to make sure both teams understand their roles and responsibilities and respect each other’s needs.

Maintenance staff need to control the security infrastructure regularly for integrity. The Owner should ideally opt for local maintenance service providers to minimise the cost of maintenance and keep response times low. This means that further emphasis should be placed on training and skills required for working at height.

Normally C&I solar PV systems are situated next to other third-party activities. This can entail extra considerations that need to be made:

  • The risk assessment should analyse dangers arising from proximity to third parties and plan countermeasures
  • O&M service providers should propose a “stakeholder training” for people working next to the installation
  • Dangerous areas should be marked in a way that is also understood by third-party personnel
INCIDENTCLASSIFICATIONCOMMENT
Inverter alarmsMinimum requirementAlarms generated by the inverter should be acknowledged at least daily. The personnel responsible for maintenance should take necessary actions within 2 days for smaller C&I installations. Faults in larger installations with central inverters need to be dealt with quicker.
Monitoring FailureMinimum requirementRemote diagnosis of monitoring failure should be completed within 2 days for smaller C&I installations. Faults in larger installations with central inverters need to be dealt with quicker. As monitoring failure is often caused by inverter failures or DC issues, this diagnosis must be done quickly to determine if the failure is limited to monitoring or if yield production is impacted. 

O&M service provider should have good guidelines and troubleshooting guides that allow the Owner to self-diagnose and resolve.

Resolution of monitoring failure without yield losses: Within 2 days for smaller C&I installations. Faults in larger installations with central inverters need to be dealt with quicker.
Inverter failureMinimum requirementAs soon as inverter failure is indicated by inverter alarms or monitoring failure a replacement or repair should be done within 1 day.
System Level Performance AlertsBest PracticeDuration and frequency of reporting should be according to the expected accuracy and availability of live irradiation data.

Best Practice is a monthly comparison, and annually as a minimum.
Module- String/Inverter Level AlertsRecommendedFor commercial projects with more than one inverter, reporting should be at the inverter level as a minimum.

String or MPPT level reporting to enable string failure alerts, is recommended where possible.
Module cleaning (and pyranometers or sensor cells if present)RecommendedThe expectation for module cleaning planning should be based on the site, installation type, size, and environmental conditions.

Actual planning of module cleaning can be adjusted based on the performance (EPI) of the system over time.

TABLE 13 - INCIDENTS COVERED BY O&M SERVICE AGREEMENTS FOR DISTRIBUTES SOLAR SYSTEM

13.1.3. Spare parts management

If economically feasible, the O&M service provider should have basic spare parts in stock. Failing this, care should be taken to select component manufacturers which can provide local service and fast replacement of faulty goods in Europe.

The inverter is the most important spare part as energy production and most monitoring processes rely on it.

13.2. Distributed residential solar portfolios

Distributed solar portfolios refer to portfolios comprising multiple, small assets installed on residential rooftops.

Ownership of assets varies from country to country and is based on the bilateral agreement between the constructor/ operator and the roof owner. Generally, there are three kinds of owners:

  1. Homeowners that own the installations on their homes
  2. Third-party companies that own the installations and usually lease the rooftop or sell the electricity produced to the owner of the rooftop at a discounted price from the one offered by utilities one
  3. Local councils or private and social housing associations that have equipped their properties with solar panels

Homeowners that own the installations on their homes have paid for the installation themselves and usually have a bilateral net- metering agreement with the local utility for the energy produced.

In the case of third parties that have paid for the installation themselves, they usually undertake the maintenance as well. The financial model depends on the bilateral agreement between themselves and the rooftop owners. Common practices include leasing the rooftop area and taking advantage of all the generated power, or selling the power produced at a discounted price to the rooftop owner.

Apart from the general aspects of rooftop solar systems, main challenges of large distributed solar portfolios are:

  • The multitude of assets: portfolios of 10,000+ installations are common
  • The variety of conditions (for example, shading, inclination, orientation, etc.)
  • The variety of equipment used: multiple inverter brands (including monitoring systems) and panels
  • The common presence of stakeholders who are not solar professionals
  • Getting access to the house for maintenance activities requires making appointments with the tenants.

13.2.1. Operations

Since physical site inspections and callouts at multiple sites mean higher costs, it is economically cheaper to invest into monitoring hardware (temperature / irradiance) on top of inverter monitoring, and implement automatic root cause analyses, where this is possible. Therefore, monitoring equipment accounts for a greater percentage of the total investment.

For large portfolios of small installations extra monitoring hardware might be too expensive. Automated analysis methodologies comparing neighbouring installations can be used in combination with irradiation data coming from meteorological stations and satellites, or theoretical clear-sky irradiation data.

Monitoring of a large portfolio of residential installations requires a different approach to monitoring an individual installation. For the latter, the inverter built-in monitoring system via Wi-Fi might be sufficient, making the tenant responsible for communication with the server.

When performing long-term monitoring of a high number of installations, using a communication channel independent from the house Internet connection, i.e., cellular communication is advised. This largely decreases the number of support calls and local interventions to resolve communications issues. It also decreases the installation cost (cabling, configuration) and the risk of cyber security issues.

For local data acquisition, three approaches can be followed:

  • Inverter manufacturer built-in system: This is often free-of-charge including access to a portal for the installer and end-user. The disadvantage is that, when multiple inverter brands are used, different monitoring systems need to be managed which makes it more complex and time consuming
  • Independent data logger: These are compatible with multiple inverter brands decreasing the dependency on a single manufacturer. The disadvantage is the extra investment
  • External energy meter: These are easy to install and often have an integrated communication module. It is the only solution when a calibrated measurement is required following the European Measurement Instrument Directive (MiD). The disadvantage is the extra investment and that only AC parameters are measured

In case only an energy meter is used in the monitoring systems, the following parameters need to be measured at the minimum:

  • AC Energy production: This is the basis for yield calculations. A resolution of minimum 15 minutes is advised for further intra-day performance analysis. In some contractual models a calibrated measurement is required following the European Measurement Instrument Directive (MiD)
  • AC voltage: In areas with a lot of local production, AC voltage can rise to a level that sends the inverter into safety mode. The level is dependent on local legislation

In case more detailed inverter data is acquired, the following parameters provide useful information:

  • Inverter alarms
  • Inverter temperature: This can give an indication of an upcoming problem or clogged ventilation holes

When monitoring large portfolios of solar PV installations, the following challenges can occur:

  • High volumes of different installations with very different characteristics
  • Base parameters (Wp, orientation, tilt) are often incorrect or missing in the monitoring database
  • Shading effects (trees, chimneys, etc.) which are season dependent resulting in errors in yield analysis
  • Local irradiation measurement is too expensive
  • Errors in yield analysis due to clipping effects  

The following best practices should be adopted:

  • Apply robust procedures during installation to start from correct parameters. Installer technicians need to provide the correct information as part of the commissioning process
  • Avoid a high variety of data acquisition methods and monitoring systems
  • Apply performance index calculations that are immune from the effects of shading (e.g., part of the day, clear sky index)
  • Compare with a pool of nearby installations to neutralise temperature, wind, and pollution effects on performance indexes

13.2.2. Maintenance

The Installer should not state that solar systems are self-cleaning and do not require any maintenance. As best practice, the Installer should educate their clients about the necessity and benefits of a regular, high-quality O&M practice for the lifetime of their solar assets. This should include a minimum yearly inspection, and cleaning and maintenance based on the environmental conditions. This will ensure the continuous safe operation of the asset and minimise H&S risks to building users. It will also maximise the energy production capability of their asset throughout its lifetime.

Preventive Maintenance of large residential portfolios is often limited to cleaning as part of a maintenance contract. Cleaning should be condition-based, rather than conform to a regular schedule. This can be combined with visual inspection of the cabling and cleaning of inverter ventilators.

It needs to be clarified to homeowners or tenants that they should not clean the panels themselves using high pressure systems. This would void the warranty.

In areas with a high density of residential solar PV installations, collective drone inspection should be considered. In a short period, thermographic data of lots of installations can be collected.

Corrective Maintenance of large residential portfolios relies heavily on a good monitoring system. Besides detecting and communicating alarms it should be able to detect decreasing performance trends.

Once an anomaly is detected, a trade-off will be made between speed of intervention and financial loss. Often it is cheaper to group interventions in a certain geographical area. A limiting factor is also the access to the house. Appointments must be made with the occupants, which can take time.

To avoid the cost of sending an intervention team on-site, tenants are often requested to perform certain actions such as removing dust from ventilators and resetting an installation (switch off/on). O&M service providers should propose training for these tasks.

More advanced residential monitoring systems calculate trends in decreasing performance and increasing inverter temperature. Both parameters predict an upcoming failure.