{"id":4338,"date":"2023-07-04T14:52:30","date_gmt":"2023-07-04T09:22:30","guid":{"rendered":"https:\/\/smarthelio.com\/?p=4338"},"modified":"2023-09-12T16:42:07","modified_gmt":"2023-09-12T11:12:07","slug":"predictive-analytics-for-solar-assets-maintenance","status":"publish","type":"post","link":"https:\/\/smarthelio.com\/predictive-analytics-for-solar-assets-maintenance\/","title":{"rendered":"Demystifying predictive analytics and preventive maintenance in the solar industry"},"content":{"rendered":"

[vc_row][vc_column column_width_percent=”80″ gutter_size=”3″ overlay_alpha=”50″ shift_x=”0″ shift_y=”0″ shift_y_down=”0″ z_index=”0″ medium_width=”0″ mobile_width=”0″ uncode_shortcode_id=”822964″][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”378482″ text_color_type=”uncode-palette”]<\/p>\n

“The future depends on what you do today.” – Mahatma Gandhi<\/span><\/span><\/strong><\/h2>\n

 <\/p>\n

T<\/span>he<\/span> rise of predictive analytics and preventive maintenance in the solar industry has gained significant <\/span>attention in recent years<\/span>. Indeed, i<\/span>n a rapidly expanding industry, with the US solar market projected to grow by 20% and Europe’s solar capacity estimated to double by 2030, the urgent need for digitalization and predictive analytics becomes <\/span>evident<\/span> to effectively manage and ensure <\/span>optimal<\/span> performance of solar installations at scale.<\/span> <\/span><\/strong><\/p>\n

However, amidst th<\/span>is constantly increasing <\/span>buzz<\/span><\/span>,<\/span><\/span><\/span> which <\/span>recently <\/span>reach<\/span>ed<\/span> a new <\/span>high<\/span> at Intersolar <\/span>Europe <\/span>2023<\/span>, there are misconceptions that <\/span>should <\/span>be addressed to <\/span>truly understand<\/span> the essence of these terms<\/span> and ensure we are all talking about the same thing<\/span>. <\/span><\/span>At<\/span> SmartHeli<\/span>o<\/span>, we are committed to<\/span> providin<\/span>g<\/span> accurat<\/span>e<\/span> and actionable insights to help solar industry stakeholders de-risk investments, maximize performance, and reduce costs.<\/span> Staying true to o<\/span>ur mission, i<\/span>n<\/span> this article we<\/span> will demystify predictive analytics and preventive maintenance and shed light on t<\/span>heir true <\/span>meaning.<\/span><\/span>\u00a0<\/span><\/p>\n

<\/h2>\n

<\/h2>\n

Understanding Predictive Analytics<\/span><\/span>\u00a0<\/span><\/strong><\/h2>\n

Predictive analytics in the solar industry <\/span>is <\/span>not only <\/span>leveraging<\/span> historical <\/span>labelled<\/span> data and advanced algorithms to forecast <\/span>production<\/span> and predict <\/span>major <\/span>component<\/span> failures like <\/span>inverters<\/span>. <\/span>It also includes the understanding of the early signs of <\/span>underp<\/span>erformance<\/span> based on<\/span> system<\/span> behaviour<\/span> anomalies.<\/span> Predictive analytics encompasses the analysis of a fault<\/span>\u2019<\/span>s <\/span>behaviour<\/span> with time<\/span> and projected losses associated with it which helps in<\/span> prioritizing <\/span>the <\/span>proactive actions to be ta<\/span>ken.<\/span> <\/strong><\/span><\/p>\n

By <\/span>analysing<\/span> past data, <\/span>understanding the real-time <\/span>electrical<\/span> patterns <\/span>in<\/span> combination with<\/span> weather patterns, solar irradiation, and system parameters, predictive analytics <\/strong><\/span>can <\/span>enable<\/span> the identification of potential issues before they cause significant <\/span>performance loss and <\/span>possible equipment<\/span> failures<\/span><\/strong>. <\/span>SmartHelio<\/span>\u2019s<\/span> Autopilot platform <\/span>uses<\/span> physics-informed AI<\/span>, <\/span>combining<\/span> M<\/span>achine <\/span>L<\/span>earning techniques with domain <\/span>expertise<\/span>,<\/span> to deliver reliable and <\/span>accurate<\/span> predictions and insights.<\/span><\/span>\u00a0<\/span><\/p>\n

\n

\u201cWhereas physics-based models are very good on post failure analysis, with more Machine Learning knowledge in our data warehouse we will then be able to recognise very small deviations at an early stage to have prediction on where your next fault may probably happen\u201d<\/span><\/i><\/p>\n

\u2013 David Ebner, Project Engineer at VERBUND<\/strong><\/p>\n<\/blockquote>\n

\"\"<\/a>
Our algorithm looks for RISO indication during humid\/rainy events and its trend\u200b in order to then calculate the risk of component failures, such as string disconnection and inverter shutdown<\/figcaption><\/figure>\n

Relying solely on AI for predictive analytics in solar, without considering physics-informed models, falls short in <\/span>several<\/span> ways. Depending on historical data alone neglects real-time changes, emerging trends, and unforeseen scenarios.<\/strong> It can lead to incomplete understanding<\/span>s<\/span>, biased predictions, and inaccurate results. By integrating physics principles into AI models, <\/span>a more comprehensive and reliable approach to predictive analytics in solar<\/span> is achieved<\/span>, <\/span>debunking <\/span>other solutions <\/span>who solely rely on AI <\/span>whilst neglecting consideration for the <\/span>underlying physics<\/span>.<\/span><\/span>\u00a0<\/span><\/p>\n

\n

“When trying to overcome these barriers in today\u2019s AI methods researchers turn back to an old principle of technology development \u2013 understanding nature and learning from it. They ask, how can we build AI systems that can tap into the knowledge we have about the physics of our world?”\u00a0<\/span><\/i><\/p>\n

\u2013 Olympia Brikis, Head of Physics-Informed AI at Siemens,<\/strong><\/p>\n

‘The hidden Potential of Physics-informed AI’ 2022 Siemens<\/em><\/p>\n<\/blockquote>\n

<\/h2>\n

The Essence of Preventive Maintenance<\/span><\/span>\u00a0<\/span><\/strong><\/h2>\n

Preventive maintenance is a critical aspect of ensuring <\/span>optimal<\/span> performance and longevity of solar installations. It involves taking <\/span>proactive<\/span> measures to prevent equipment failures and mitigate potential risks<\/span> of underperformance<\/span>. <\/span>P<\/span>reventive maintenance <\/span>today <\/span>focuses <\/span>mainly <\/span>on<\/span> regular inspections, servicing, and repairs<\/span> of the mechanical and electrical infrastructures of a PV plant<\/span>. <\/span><\/span>However, these regular inspections <\/strong><\/span>are often periodic based<\/span>, not issue focused, <\/span>or <\/span>can be <\/span>investigative in a broad sense <\/span>only after <\/span>a failure <\/span>or underperformance <\/span>has <\/span>been <\/span>identified<\/span>. <\/span><\/strong><\/span><\/p>\n

Lacking data-driven intelligence, such interventions are <\/span>not <\/span>an<\/span> efficient but reactive use of resources<\/span>, <\/span>and they come at significant cost<\/span>. <\/span><\/strong>For example, <\/span>preventive <\/span>maintenance based solely on regular interventions <\/span>cannot consider<\/span> data-driven <\/span>intelligence <\/span>that may <\/span>detect <\/span>component<\/span> behaviour<\/span> anomalies<\/span>. By incorporating real-time data<\/span> analytics, predictive maintenance can <\/span>opti<\/span>mize<\/span> per<\/span>formance<\/span> by <\/span>iden<\/span>tifying<\/span> and<\/span><\/strong> addressing potential issues before they escalate<\/strong>, <\/span>ensuring <\/span>optimised<\/span><\/strong><\/span> system performance and reducing unnecessary maintenance activities<\/span> and cost<\/span>.<\/span><\/span><\/strong>\u00a0<\/span>[\/vc_column_text][\/vc_column][\/vc_row][vc_row unlock_row=”” row_height_percent=”0″ back_color=”color-lxmt” overlay_color=”color-xsdn” overlay_alpha=”50″ equal_height=”yes” gutter_size=”0″ column_width_percent=”86″ preserve_border=”yes” preserve_border_tablet=”yes” preserve_border_mobile=”yes” border_color=”transparent” shift_y=”0″ z_index=”0″ uncode_shortcode_id=”208379″ back_color_type=”uncode-palette” overlay_color_type=”uncode-palette” border_color_type=”uncode-palette”][vc_column width=”1\/2″][vc_custom_heading heading_semantic=”h3″ text_size=”h1″ sub_reduced=”yes” uncode_shortcode_id=”928080″]<\/p>\n

How i<\/span>s<\/span> it working?<\/span><\/strong><\/h2>\n

[\/vc_custom_heading][vc_column_text uncode_shortcode_id=”194983″]In solar preventive maintenance, we can draw a parallel to tachycardia,<\/strong> a condition that signals an impending heart attack. Just as specific heart frequency patterns indicate an issue, regardless of who the person is, anomalies in solar component behaviour, such as inverter shutdown or tracker malfunction, can be detected early.<\/p>\n

By recognizing these patterns, proactive measures can be taken to prevent system failures and optimize solar plant performance.<\/strong> Similar to understanding the biology of the heart to be able to detect a heart attack, one must deeply understand the physics of solar to make sense of the electrical patterns.[\/vc_column_text][\/vc_column][vc_column width=”1\/2″][vc_single_image media=”4371″ media_width_percent=”100″ uncode_shortcode_id=”330107″][\/vc_column][\/vc_row][vc_row unlock_row=”” row_height_percent=”5″ overlay_alpha=”50″ gutter_size=”3″ column_width_percent=”100″ shift_y=”0″ z_index=”0″ uncode_shortcode_id=”154646″][vc_column width=”1\/1″][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”980745″ text_color_type=”uncode-palette”]By <\/span>identifying<\/span> and addressing issues at an early stage, preventive maintenance minimises downtime, improves energy production, and reduces maintenance costs. <\/span>SmartHelio<\/span> adopts a comprehensive preventive maintenance strategy that goes beyond relying solely on SCADA<\/span>\/<\/span>monitoring <\/span>alerts. We proactively assess equipment health, <\/span>provide<\/span> recommendations based on data-driven insights and prompt preventive maintenance scheduling based on <\/span>component<\/span> information analytics developed over<\/span> 5GW <\/span>of global assets management.<\/span><\/span>\u00a0<\/span><\/strong><\/p>\n

<\/h2>\n

<\/h2>\n

Monitoring vs. Analytics in Preventive Maintenance<\/span><\/span>\u00a0<\/span><\/strong><\/h2>\n

It is important to understand the distinction between monitoring and analytics in the context of preventive maintenance. Monitoring involves real-time data collection and basic alerts<\/span> that<\/span> provid<\/span>e<\/span> immediate information about the current state of the system. Whil<\/span>st<\/span> monitoring is essential for detecting immediate issues, it lacks the in-depth analysis and long-term insights that analytics offer. <\/span><\/span><\/p>\n

Analytics<\/strong>, on the other hand, involves advanced data processing, correlation analysis, and predictive <\/strong><\/span>modelling<\/span> in order to<\/span> provide<\/span> technical, <\/span>financial<\/span> and economical actionable intelligence<\/span><\/strong>.<\/span> It uncovers hidden patterns, <\/span>identifies<\/span> trends, and <\/span>provides<\/span> actionable insights for preventive maintenance planning. <\/span>SmartHelio’s<\/span> Autopilot<\/span> platform <\/span>empowers<\/span> monitoring<\/span> systems<\/span>, <\/span>enabling granular analysis, <\/span>accurate<\/span> fault predictions, and data-driven decision-making for efficient preventive maintenance strategies. The <\/strong><\/span>plug-in <\/span>software <\/span>platform <\/span>integr<\/span>ates <\/span>seamlessly<\/span> with <\/span>the<\/span> existing <\/span>SCADA or CMS<\/span> monitoring<\/span> system<\/span> in order to<\/span> automate <\/span>and <\/span>accelerate <\/span>accurate<\/span> fault classification, failure prediction and action prescription.<\/span><\/strong><\/span>\u00a0<\/span><\/strong><\/p>\n

\u201cIncrease PV power generated by the panel is one value. There was a series of ground fault, which may cause over-voltage and even fire risk [\u2026] Safety is number one priority for Schneider. We see safety improvement as a second value to (SmartHelio\u2019s) technology\u201d<\/i><\/p>\n

\u2013 Luc Hossenlopp, CTO Digitalisation at Schneider Electric<\/strong><\/p><\/blockquote>\n

[\/vc_column_text][vc_empty_space empty_h=”1″][vc_column_text uncode_shortcode_id=”952088″]<\/p>\n

Differentiating <\/span>SmartHelio<\/span> as a Reference in Predictive Analytics<\/span><\/span>\u00a0<\/span><\/strong><\/h2>\n

To <\/span>truly understand<\/span> the meaning <\/span>SmartHelio<\/span> attributes to predictive analytics<\/span><\/span>,<\/span><\/span><\/span> let’s explore concrete <\/span>use<\/span> cases <\/span>in action<\/span><\/span><\/h3>\n

[\/vc_column_text][\/vc_column][\/vc_row][vc_row row_height_percent=”0″ back_color=”color-lxmt” overlay_alpha=”50″ gutter_size=”3″ column_width_percent=”100″ shift_y=”0″ z_index=”0″ uncode_shortcode_id=”183700″ back_color_type=”uncode-palette”][vc_column width=”1\/3″][vc_custom_heading text_color=”color-487870″ uncode_shortcode_id=”159004″ text_color_type=”uncode-palette”]<\/p>\n

Early Detection of Declining Panel Efficiency<\/span><\/span><\/strong><\/h3>\n

[\/vc_custom_heading][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”170486″ text_color_type=”uncode-palette”]By <\/span>analysing<\/span> historical data and tracking performance trends, <\/span>SmartHelio’s<\/span> predictive analytics algorithms c<\/span>ould<\/span> identify<\/span> gradual <\/span>reduction<\/span>s <\/span><\/strong>in panel efficiency<\/strong>.<\/span> In our latest finding from a <\/span>75MW site in Italy, <\/span><\/strong>our software<\/span> detected 0.78% higher degradation at the end of 3<\/span><\/span>rd<\/span><\/span> year<\/span> since <\/span>commissioning.<\/span> <\/span><\/p>\n

This early detection enable<\/span>d<\/span> timely<\/span> interventions<\/span> a<\/span>nd warranty claim coordination with the OEM<\/span><\/strong> to ensure <\/span>optimal<\/span> performance and maximize energy production.<\/span><\/span>[\/vc_column_text][\/vc_column][vc_column width=”1\/3″][vc_custom_heading text_color=”color-487870″ uncode_shortcode_id=”136367″ text_color_type=”uncode-palette”]<\/p>\n

Predicting Potential Faults in the Inverter System<\/span><\/span><\/strong><\/h3>\n

[\/vc_custom_heading][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”209787″ text_color_type=”uncode-palette”]Leveraging<\/span> physics-informed ML models, <\/span>SmartHelio’s<\/span> analytics platform <\/span>identifie<\/span>d<\/span> patterns and anomalies in <\/span><\/strong>inverter system data<\/strong>. Overheating issues were <\/span>detected<\/span> as the probable cause. Our predictive intervention avoided <\/span>significant resources<\/span> to find the root cause & fix the failure<\/strong><\/span> of this 1.75MW inverter<\/span><\/strong>.<\/strong> <\/span><\/span><\/p>\n

By detecting potential faults before they occur, our platform <\/span>enabled <\/span>proactive<\/span>, targeted<\/span> maintenance, reducing <\/span>downtime<\/span> and avoiding costly system failure<\/span><\/span> losses<\/span> (~EUR 787 per day)<\/span>.<\/span><\/strong><\/span>\u00a0<\/span><\/strong>[\/vc_column_text][\/vc_column][vc_column width=”1\/3″][vc_custom_heading text_color=”color-487870″ uncode_shortcode_id=”110773″ text_color_type=”uncode-palette”]<\/p>\n

Optimising Trimming Schedules Based on Vegetation Rate Projection<\/strong><\/h3>\n

[\/vc_custom_heading][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”830830″ text_color_type=”uncode-palette”]SmartHelio’s advanced analytics considered utilized advanced pattern recognition algorithms to detect evolution of horizons profile. Supervised ML models then identified pattern of vegetation<\/strong>, and optimized trimming interventions based on projected vegetation growth trends which was forecasted to be impact the performance by 0.018% per week in this 20MW plant.<\/strong> <\/span><\/p>\n

By intelligently planning trimming operations, solar assets could maintain optimal performance while minimizing unnecessary trimming costs.<\/span>\u00a0<\/span>[\/vc_column_text][\/vc_column][\/vc_row][vc_row row_height_percent=”0″ overlay_alpha=”50″ gutter_size=”3″ column_width_percent=”100″ shift_y=”0″ z_index=”0″ uncode_shortcode_id=”825340″][vc_column width=”1\/1″][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”178326″ text_color_type=”uncode-palette”]<\/p>\n

These examples highlight <\/span>SmartHelio’s<\/span> ability to deliver <\/span>accurate<\/span> and actionable insights, empowering solar industry stakeholders to make informed decisions and <\/span>optimise<\/span> their operations.<\/span><\/span>\u00a0<\/span><\/strong><\/h4>\n

<\/h2>\n

 <\/p>\n

Conclusion<\/span><\/span><\/strong><\/h2>\n

U<\/span>nderstanding the true meaning of predictive analytics and preventive maintenance is crucial <\/span>for<\/span> today’s solar industry. <\/span>Levera<\/span>ging its value <\/span>will <\/span>enable solar industry stakeholders to de-risk investments, maximise <\/span>performance, and reduce costs. <\/span><\/strong>SmartHelio’s<\/span> commitment to <\/span>providing<\/span> accurate<\/span> and actionable insights sets us apart as a benchmark in the industry. <\/span>Through early fault detection, optimised maintenance strategies, and reliable predictions, <\/span>SmartHelio<\/span> empowers decision-makers with data-driven approaches for a sustainable and efficient solar future.<\/span><\/span>\u00a0<\/span>[\/vc_column_text][\/vc_column][\/vc_row]<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

[vc_row][vc_column column_width_percent=”80″ gutter_size=”3″ overlay_alpha=”50″ shift_x=”0″ shift_y=”0″ shift_y_down=”0″ z_index=”0″ medium_width=”0″ mobile_width=”0″ uncode_shortcode_id=”822964″][vc_column_text text_color=”color-nhtu” uncode_shortcode_id=”378482″ text_color_type=”uncode-palette”] “The future depends on what you […]<\/p>\n","protected":false},"author":7,"featured_media":4389,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_seopress_robots_primary_cat":"none","_seopress_titles_title":"","_seopress_titles_desc":"What is predictive maintenance and why only AI is not the solution to underperforming solar assets. Discover predictive analytics.","_seopress_robots_index":"","footnotes":"","_links_to":"","_links_to_target":""},"categories":[4],"tags":[46,49,47,40,51,52],"_links":{"self":[{"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/posts\/4338"}],"collection":[{"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/comments?post=4338"}],"version-history":[{"count":1,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/posts\/4338\/revisions"}],"predecessor-version":[{"id":5656,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/posts\/4338\/revisions\/5656"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/media\/4389"}],"wp:attachment":[{"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/media?parent=4338"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/categories?post=4338"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/smarthelio.com\/wp-json\/wp\/v2\/tags?post=4338"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}