CONTACT

What factors are important in O&M for PV power plants?

Ulrike Jahn, TÜV Rheinland

Ulrike Jahn: "Error avoidance begins in the planning stage". TÜV Rheinland’s Ulrike Jahn responds to seven questions on the operation and maintenance of solar parks.


Ms. Jahn, you work on operation and maintenance (O&M) for PV power plants. What does that entail?

First of all, it’s about assuring the quality of PV components in the field using methods and processes for adequate error detection, particularly in large installations. Secondly, it’s about quantifying technical risks before and during operation. We carried out an EU project which involved not just assessing the risks, but also developing a method (CPN) for costing them. Thirdly, it’s about recommending suitable measures for operation and maintenance – not just for the mild European climate, but for other regions, too.

 

What factors are important in O&M for solar parks?

Quality assurance during every stage of O&M services, from planning, installation and operation to any repowering processes. Unfortunately, we often observe serious installation errors in the field, which are partly caused by insufficiently trained staff. We are struggling with intense price pressure and a lack of standardization concerning error identification and measures. Minimum requirements must be implemented even during the planning phase in order to avoid poor O&M management and a lack of coordination of the various service providers. Professional, timely data analysis is also important in order to identify any errors at an early stage. We like to talk about the “ten times ten times ten” rule, which means that dealing with an error costs ten times more during installation than during planning, and ten times more again during operation and maintenance. In other words, if an error can be identified and avoided in the planning phase, it will be a thousand times cheaper than if the error is only noticed and rectified during maintenance work.

 

So avoiding errors is better than detecting them early on. Can the risk of certain kinds of damage be estimated?

It is essential to estimate and assess the true risks – from the planning phase onwards. For example, in a tropical location with high humidity levels, it’s important to think about the risk of potential-induced degradation (PID) of the modules. The specific risk of PID when using the chosen module type in that location must be assessed in advance, and the costs of minimizing this risk by performing prior lab tests must be considered. Another preventative measure is adequate spare parts management, which ensures that a sufficient number of spare parts are kept in stock for short-notice use on the site. And there is another, even more important risk management measure: training the O&M team. The invitation to tender should specify that training is to be provided. TÜV Rheinland offers a service certificate for this.

 

What are the typical errors that occur in PV power plants and how can they be avoided?

I’ll mention four. PID in the field remains a serious problem. One way of preventing it is carrying out laboratory PID tests using small random samples. Such tests cost money, of course, but these costs are relatively low compared with the costs of any actual damage in the field. Another option is using drones for quick, large-scale infrared analysis. A drone can cover a solar field with a capacity of one megawatt in an hour, and automatically evaluate the data.

A second common issue is chalking, which is when the rear side of a module becomes bleached or lined, potentially causing material damage and subsequent corrosion due to humidity entering the module. This is a material issue which has to do with the polymer layers of the back sheet. It is possible to accurately predict the risk of chalking by exposing the rear of a module to UV radiation in a climatic chamber and running temperature cycles.

 

And what about the other two errors?

The third point is problems with connectors – both within the module and when connecting cables. Experience has shown that connectors have a high potential for error. They may have been connected incorrectly, or the modules may not even have been connected to form a string. Installation errors like these can generally be identified through infrared analysis to show where entire strings have not been connected or connectors from different manufacturers have been used. More attention could be paid to components during the planning phase, to avoid connectors from different manufacturers being used.

The fourth issue I would like to mention concerns soiling of the modules, which can be caused by sand, dust, snow or bird feces. Suitable cleaning strategies should be created during the planning phase. What is suitable depends very much on the location, so the soiling rate on the site should be measured to establish an adequate cleaning strategy during the planning phase. Weekly automatic cleaning may well be necessary. This makes it crucial to know the exact spacing between the module rows, so that the cleaning vehicle does not damage or scratch the modules.

 

Digitalization brings with it new possibilities for data processing. How has this changed the way that PV power plants are monitored?

I think this is a growing trend. I remember that two or three years ago, we still had problems identifying modules recorded by drones. Data processing and better imaging have improved this process a lot, making it much more accurate. But we can still expect new developments based on intelligent monitoring to reduce costs.

 

What will it take for service providers to succeed in a growing O&M market in the future?

According to estimates, artificial intelligence could drive energy yields up by five percent, and cut O&M costs by as much as ten percent. These are significant figures which will also have an effect on electricity generation costs (LCOE). We always need to strike the right balance between maximum quality assurance and affordable costs. For most measures, the expense necessary to go from ninety-five percent to one hundred percent certainty is disproportionately high. What’s important is standardization and automatic data diagnosis. The diagnosis process will include a type of decision-making function in the future – this will shorten downtimes and reduce the long-term generation costs for PV power plants.

 

To learn more on this topic, attend the session "Best Practice Concepts for Operation & Maintenance of PV Power Plants" on May 14, 2019 at the Intersolar Europe Conference.