Chapter 8: Monitoring, Alarming, Operation and Maintenance, Cleaning and SLA Hardcore Version

8.1 Monitoring is not about how much power is generated, but about whether the system deviates from expectations.

Many teams regard the monitoring platform as a “power generation digital display board” for customers, but the idea of ​​​​IEC 61724 is closer to ‘performance monitoring and analysis’. It emphasizes not only collecting data, but also judging whether the system performs as expected through parameters such as irradiation, array output, system output, and temperature. In other words, the purpose of monitoring is not to look at a beautiful number every day, but to identify deviations.

For household and small commercial projects, although it is not necessary to deploy large-scale power station-level instruments, the same thinking must be formed: at least it must be able to compare the power generation performance under similar weather conditions today, last week, last month and history; at least it must be able to identify problems such as offline, power generation sudden drop, repeated inverter alarms, single-channel MPPT abnormalities, communication dropouts, etc. If monitoring is only turned on after a customer complains, it is not called monitoring, it is just called after-the-fact review.

Therefore, monitoring should be defined as a ‘tool for detecting deviations’ rather than a ‘page that displays results’. This will directly affect the way the after-sales team looks at data every day.

8.2 What are the minimum types of indicators that should be included in the monitoring dashboard?

对中小型屋顶项目，不需要一开始就复制电站级 SCADA，但监控看板至少应包含五类信息：Availability、Daily/weekly/monthly power generation、Inverter and MPPT status、Alarm record、Communication online status. It would be more valuable if historical comparisons and weather comparisons could be further added.

One of the most overlooked is availability and communication presence. Many project customers think the system is fine when they see numbers on the monitoring page, but in fact the system may have been offline for several days, but the inverter is still running locally; or the communication is online, but a certain MPPT channel is abnormal for a long time and is covered up by the accumulated power generation. The after-sales team must learn to look at both the total quantity and the structure.

一个实用培训方法是：不要只让新人看‘今天发了几度’，而要问他‘为什么和上周同类天气不一样、哪一路变化最大、是不是监控问题还是电气问题’。 By asking these questions, operation and maintenance capabilities begin to truly form.

8.3 How to judge ‘low power generation’, don’t just rely on feeling

The most dangerous habit in operation and maintenance is to say "there seems to be less hair today" based on naked eyes and experience. A better method is to establish a set of simple deviation judgment rules. PNNL's public operation and maintenance best practices mention that if the system output deviates from the expected range of average solar irradiation by about ±10%, troubleshooting should begin; at the same time, dirt is one of the most common O&M problems. In other words, low power generation is not necessarily a fault, it can also be caused by dirt, obstruction, weather deviation or communication problems.

For small and medium-sized roof projects, it is recommended to establish at least three comparison dimensions: comparison with history in the same month, comparison with adjacent sunny days, and comparison with similar projects in the same installation area. If all three show abnormality, then enter on-site or remote diagnosis. This is more economical and more professional than "coming to the door as soon as the customer says less".

We also need to explain the dirty logic clearly. Dirt loss is not always linear, nor is dirtier the easier it is to see from a distance. NREL/NLR's public research emphasizes the concept of soiling ratio, or annual soiling loss, indicating that the impact of soiling needs to be judged through data, not just visual inspection.

8.4 Cleaning is not a cleaning action, but a cost-benefit decision

It is easiest for customers to understand cleaning as "washing when it is dirty", and it is also easy for companies to understand cleaning as "door-to-door service". In fact, whether cleaning is worth doing, how often to do it, and how to do it are all cost-benefit issues. Public information from PNNL points out that the cleaning method and frequency depend on the degree of dirtiness, system size, site environment and labor/equipment costs; strategies may be completely different in urban and low-dust environments and near dirt roads, agricultural areas, and seaside environments.

Therefore, training should not teach newcomers to "clean it every few months", but teach them to judge whether the current site is dirty enough to affect revenue, whether the cleaning cost is lower than the expected additional revenue, and whether customers care more about power generation or appearance. For high-net-worth households, cleaning may also have aesthetic value; for small businesses, it is more about income and stability.

A mature after-sales system should be able to give customers at least three suggestions: no cleaning for now, regular cleaning recommended, and focused inspection and cleaning recommended. In this way, what customers feel is professional judgment, rather than a visit just for the sake of visiting.

8.5 Periodic inspection and IEC 62446: Why operation and maintenance should have annual physical examination thinking

The public summary of IEC 62446-1 applies not only to initial delivery but also to subsequent re-inspections and maintenance. For operation and maintenance, this means that projects should not only be ‘installed and tested once’, but should have a periodic inspection mentality. Perform appropriate revalidation at least annually or as required by the AC system to check that the equipment is still in safe and correct operating condition.

This type of periodic inspection can be a lightweight version for household projects: appearance, fastening, visible cables, inverter alarms, online monitoring, power generation deviations, markings and customer feedback; for small commercial projects, more systematic electrical inspections and document updates should be added appropriately. As long as the team establishes the awareness of ‘annual physical examination’, many faults will be discovered before customers complain.

It needs to be clear: maintenance is not about fixing something that is broken, but about extending the period during which the system remains in a correct state.

8.6 SLA, work orders and customer experience: the real management language of after-sales

After-sales is not managed by empty words such as ‘deal with it as soon as possible’, but by SLA and work orders. The so-called SLA does not necessarily have to be very complicated at the beginning, but at least it must distinguish between response time limit and resolution time limit. It is obvious that the same response criteria should not be used for general consultation, monitoring offline, inverter alarms, shutdown faults and safety risks.

Work orders are the vehicle for organizational learning. An after-sales team without a work order system will only solve similar problems repeatedly but cannot accumulate knowledge. Conversely, as long as each problem can be categorized as 'user misunderstanding, monitoring communication, equipment failure, installation issues, external factors', you will quickly know where the most common problems are coming from and who should be trained next.