Solar Now

10W 12V Portable Polycrystalline Solar Panel With Clip – For Charging 9-12V Batteries, Outdoor Use  ruhrkanal.news
source

Cloudy skies. A stray shower or thunderstorm is possible. Low 41F. Winds S at 10 to 15 mph..
Cloudy skies. A stray shower or thunderstorm is possible. Low 41F. Winds S at 10 to 15 mph.
Updated: April 4, 2026 @ 10:53 pm
A solar farm can be seen from Mohr Road in the town of Florida in November 2024. A solar farm being built in New Baltimore will be acquired by New York state in 2027.

A solar farm can be seen from Mohr Road in the town of Florida in November 2024. A solar farm being built in New Baltimore will be acquired by New York state in 2027.
NEW BALTIMORE — The Hannacroix Solar farm in New Baltimore will be taken over by a state entity once it is completed.
The New York Power Authority on Tuesday announced an agreement that would result in the authority helping with the development of the five-megawatt solar farm on Railroad Avenue, with a full acquisition of the farm set for 2027.
“NYPA is providing the financing needed to move Hannacroix Solar through its remaining preconstruction development work,” Power Authority spokesperson Alex Chiaravalle said Tuesday. “After completing a final round of due diligence, the Power Authority will finance the project’s construction — expected to begin later this year — and fully acquire the project upon its completion in 2027.”
The project is being developed by Teichos Energy, a Seattle-based solar energy company.
“NYPA has long been recognized as an innovator in the electric-power industry and we welcome the opportunity to advance their goals of making New York’s electric power mix cleaner while stabilizing power costs,” Teichos Energy CEO Stephen Voorhees said in a statement.
The acquisition is part of the state’s Renewable Energy Access and Community Help program, designed to provide renewable energy credits to low-income New Yorkers, according to the state’s website.
The project is also part of the authority’s renewable-energy strategic plan, which outlines the state’s efforts to develop, own and operate renewable-energy projects, the announcement reads.
The credits are funded through renewable-energy production from properties developed by or for the authority.
Gov. Kathy Hochul authorized several state departments, including the Power Authority, to fast track renewable-energy developments before federal tax credits expired.
“Since NYPA was authorized less than three years ago to support large-scale renewable-energy development, it has worked diligently to stand up a program capable of accelerating clean-energy projects across New York State,” Power Authority President and CEO Justin Driscoll said in a statement. “Amid headwinds affecting the industry, NYPA established new business structures, assembled a team of seasoned professionals, and refined a portfolio of project opportunities. In 2026, those efforts will bear fruit.”
The five-megawatt facility will be developed on nearly 54 acres of land in New Baltimore. Construction is expected to begin later this year, Chiaravalle said.
The authority is also set to help finance the project’s preconstruction development and Gordian Energy Systems as the project’s engineering, procurement, construction and contractor to build the project, according to Bill Parkhurst, Teichos vice president of marketing and finance.
“We are excited that the New York Power Authority has stepped in as a strong financing and development partner,” Parkhurst said.
The New York Power Authority is the largest state power organization in the country, operating 17 facilities and more than 1,550 miles of transmission lines.
The solar farm is expected to go online in late 2027.
{{description}}
Email notifications are only sent once a day, and only if there are new matching items.
Success! An email has been sent to with a link to confirm list signup.
Error! There was an error processing your request.
HV360 headlines to your inbox!
Get the Register Star and Daily Mail E-edition delivered directly to your inbox!
Your browser is out of date and potentially vulnerable to security risks.
We recommend switching to one of the following browsers:
Sorry, an error occurred.

Already Subscribed!

Cancel anytime
Account processing issue – the email address may already exist
Have the latest food and drink articles delivered to your inbox every Thursday.
Local and state politics made simple. Hits your inbox every Tuesday.
A weekly roundup of local crime news—delivered every Wednesday
Top stories and breaking news, delivered daily at 5:30 a.m. and 6:00 p.m.
Honoring lives in our community—new obits shared daily.
Get the day’s full paper—The Daily Gazette, Amsterdam Recorder & Leader Herald—delivered to your inbox daily by 5 a.m.
Expert race picks delivered daily during the Saratoga track season.
HV360 headlines to your inbox!
Remembering those we’ve lost. Delivered Tuesday through Saturday.
Top local stories delivered to your inbox Monday through Saturday.
Leader Herald & Amsterdam Recorder Obituaries. Remembering those we’ve lost. Delivered Monday through Saturday.
Major news, right when it happens—sent straight to your inbox.
Major news, right when it happens—sent straight to your inbox.
Get the Register Star and Daily Mail E-edition delivered directly to your inbox!
New! Spotlight News now has an e-edition.
Have the latest local news delivered every morning so you don’t miss out on updates.
Get The Spot518 e-edition in your email weekly.
The latest Spot518 headlines twice a week.
A weekly roundup of our top stories from The Spot518 —delivered every Saturday.
A weekly roundup of our top stories from across all our publications—delivered every Saturday.
Get Your Niskayuna e-edition in your email weekly.

Thank you .
Your account has been registered, and you are now logged in.
Check your email for details.
Invalid password or account does not exist
Submitting this form below will send a message to your email with a link to change your password.
An email message containing instructions on how to reset your password has been sent to the email address listed on your account.
No promotional rates found.

Secure & Encrypted
Secure transaction. Secure transaction. Cancel anytime.

Thank you.
Your gift purchase was successful! Your purchase was successful, and you are now logged in.
A receipt was sent to your email.

source

Newsfrom Japan

Tokyo, April 3 (Jiji Press)–The Japanese government on Friday adopted a bill that requires operators of mega solar power plants to submit plans on how they will dispose of used panels.
The bill is aimed at getting mega solar farm operators to cooperate on recycling used panels, in order to reduce the volume of those that go to landfill.
The country is expected to see the amount of dumped used solar panels reach up to 500,000 tons around 2040, about six times the current volume.
The bill stipulates that operators set out in their plans the volume of panels to be disposed of, along with when and how they will be dumped, among other aspects.
Operators will be asked to dispose of used panels in ways that allow them to be recycled into materials.

[Copyright The Jiji Press, Ltd.]

Jiji Press

source

Cloudy with periods of rain. Thunder possible. Low 39F. Winds W at 10 to 20 mph. Chance of rain 100%..
Cloudy with periods of rain. Thunder possible. Low 39F. Winds W at 10 to 20 mph. Chance of rain 100%.
Updated: April 4, 2026 @ 11:33 pm
Ed Sekerak, standing at left center, offers testimony regarding property values in the area surrounding the proposed Solar Flats project. The hearing Thursday lasted over four hours.

Ed Sekerak, standing at left center, offers testimony regarding property values in the area surrounding the proposed Solar Flats project. The hearing Thursday lasted over four hours.
Corry City Council convened a public conditional use hearing Thursday on a proposal by Solar Flats LLC to construct a 3-megawatt AC solar energy facility on approximately 16 acres of land along South Shady Avenue, owned by Troyer Growers Inc. The session drew a crowd of about 45, and featured hours of technical testimony, legal argument and public comment from residents who live within sight of the proposed installation.
Mayor Jeff Fike presided, joined by Councilmen William “Buzz” Hammond and William Roche and Councilwoman Charles “Chuck” Gray. Councilman Andrew Sproveri recused himself, citing concern over potential bias.
At the close of testimony, Council adjourned to executive session and announced it would reconvene at a later date to render a formal decision.
The project
Annika Schiffer-Delagard, project development associate for Greenwood Sustainable Infrastructure, described Solar Flats as a relatively modest commercial solar installation designed to feed power directly into the grid through Penelec. She said the project would have the capacity to power between 300 and 400 homes.
Schiffer-Delagard outlined what she characterized as the project’s low environmental footprint: no lighting, no signage, no smoke, no odor. She said maintenance would require only approximately three truck visits per year. She added that the panels themselves are silicon and glass, are UL-certified and pose no leaching risk to soil or groundwater.
Schiffer-Delagard said the company contracted for noise and glare studies ahead of the application. A noise analysis found all operational sound would fall at or below 45 decibels, comparable to a household dishwasher, and below ambient levels. A glint-and-glare study, performed by Fisher Associates engineer Steve Mellott, found minimal “green glare” — the lowest category, likened to sunlight reflecting off snow or a pond — at two locations, the first totaling 47 minutes of exposure annually, occurring in spring and fall near an agricultural barn at 340 Sample Flats Road; the second at the residence of Dr. Kurt Lund, totaling 16 minutes per year between May and June.
The project will include tracker panels — motors that rotate the arrays to remain perpendicular to the sun — powered by electricity generated by the project itself. Mellott testified the site plans include erosion and sediment controls, post-construction stormwater management and vegetative screening on three of four sides of the property. The north side, he said, is already bordered by existing woods.
Greenwood representative Drew Rogerson told Council that a 100-foot setback from residences is included and that the company maintains a 24/7 operations-and-maintenance team to respond to any damage, such as panels broken by hail. He also addressed concerns about site selection, assuring the audience that Corry was chosen because a willing landowner came forward and the utility interconnection proved feasible — not because the company believed the community would offer little resistance.
Legal arguments over zoning classification
Attorney Kevin Barley of Steptoe and Johnson in Pittsburgh, representing Solar Flats LLC and Greenwood Sustainable Infrastructure, argued that the application was filed at the city’s own direction. He said the city’s interim zoning officer, Samantha Vollentine, consulted with City Solicitor Lydia Caporosa in spring 2025 and instructed Solar Flats to apply under the “special residential and commercial projects” conditional use category in the R-1 single family residential district.
Barley reminded Council that a conditional use is presumed by the legislature to be consistent with the zoning plan and not inherently adverse to the public interest. He said the hearing was not about whether residents or officials favor or oppose solar energy, but whether the applicant meets the conditions set forth in the ordinance. He also noted that under the existing R-1 ordinance, uses with far greater potential impact — including mineral extraction, cemeteries, hospitals, nursing homes and funeral homes — are already permitted by right or conditional use.
Attorney Edward Betza of Elderkin Law Firm in Erie, representing objecting resident Kelly Goodsel and other South Shady Avenue neighbors by extension, offered a sharply different reading of the ordinance. He argued that the city’s zoning code, adopted in 1991 and never updated to address solar development, contains three distinct use categories: “planned commercial project,” “industrial use,” and “special residential and commercial project.”
Betza contended that a solar energy facility is plainly a planned commercial project, a category permitted only in I-1 and I-2 industrial districts, not R-1, or possibly an industrial use, which is similarly barred from the residential zone. Betza argued that the “special residential and commercial project” category requires both a residential and a commercial component, and that Solar Flats, having no residential element whatsoever, cannot legitimately claim that classification. 
“The word ‘and’ is significant,” he told Council. “I didn’t write it.”
Beyond the classification dispute, Betza argued that the applicant failed to meet its burden of proof under Section 605 of the zoning ordinance, pointing out that the testimony addressed neither fire protection nor electrical distribution disturbance, both required criteria. He also cited the ordinance’s provision that reflective materials or lighting “which produce objectionable, direct or reflected glare on adjoining properties shall not be permitted,” arguing the standard is absolute. 
“It says none — zero. Shall not be permitted,” Betza said.
Property values and decommissioning
Columbus resident Edward Sekerak, a state-certified general real estate appraiser who testified he has completed more than 10,000 appraisals over his career, said he found no evidence in available studies that a solar project of Solar Flats’ size and location would have a measurable negative effect on surrounding property values. He cited research from Virginia Tech and other institutions. Under cross-examination, however, Sekirak acknowledged he had not read the cited studies in their entirety.
On decommissioning, Schiffer-Delagard said the project’s anticipated operational life is 25 to 30 years. She said Solar Flats plans to secure a surety bond covering 110% of estimated decommissioning costs, calculated using 2026 pricing and an inflation rate. 
When asked by Councilwoman Gray whether the panels might simply be replaced and the facility continued beyond that horizon, Schiffer-Delagard said she could not commit to that scenario. 
When Mayor Fike asked how many sites Greenwood has decommissioned to date, she replied, “Zero.”
Public comment
Ten members of the public addressed Council, the large majority in opposition. Kelly Goodsel, whose property abuts the proposed installation, organized his objections into three categories: environmental concerns, including potential stormwater runoff into a nearby spring and existing wetlands to the north; property and home value impacts; and legal and ordinance compliance questions.
Goodsel also noted that property owners along Stewart Road in Concord Township, whose land would be directly affected by the project, received none of the informational mailings Greenwood sent to Shady Avenue residents in October 2025, and therefore had no advance notice of the proceedings. He said a portion of his own 20-acre property lies in Concord Township.
Lund echoed Goodsel’s testimony and raised additional concerns about glare hazards for motorists traveling roads near the project.
One speaker broke from the majority. Ed Spitman, who said he has lived near the area for 66 years, argued that the landowner has a right to use his property as he sees fit. 
“Everybody wants to make money, so everybody’s got to sacrifice a little bit,” Spitman said. “You’ve got to be open-minded to everybody.”
Planning commission previously split
The Corry Planning Commission previously voted 3–2 against recommending approval of the conditional use application, leaving the final determination to City Council. See the March 6 edition of The Journal for more information. 
Council has not announced a date for its decision.
Your comment has been submitted.

Reported
There was a problem reporting this.
Log In
Keep it Clean. Please avoid obscene, vulgar, lewd, racist or sexually-oriented language.
PLEASE TURN OFF YOUR CAPS LOCK.
Don't Threaten. Threats of harming another person will not be tolerated.
Be Truthful. Don't knowingly lie about anyone or anything.
Be Nice. No racism, sexism or any sort of -ism that is degrading to another person.
Be Proactive. Use the 'Report' link on each comment to let us know of abusive posts.
Share with Us. We'd love to hear eyewitness accounts, the history behind an article.
Sorry, there are no recent results for popular videos.
Sorry, there are no recent results for popular commented articles.
Success! An email has been sent to with a link to confirm list signup.
Error! There was an error processing your request.
Would you like to receive our daily news? Sign up today! 
Your browser is out of date and potentially vulnerable to security risks.
We recommend switching to one of the following browsers:

source

Many countries and regions are gradually facing a situation where early-installed photovoltaic modules are reaching the end of their service life; this establishes a solid market foundation for the establishment of solar module recycling plant. For enterprises planning to invest in the solar module recycling sector, selecting suitable country in which to site a facility is a critical factor for project success.
Based on the scale of the solar PV market and the associated recycling demand, the following countries are particularly worthy of consideration. These countries possess substantial installed solar capacity or rapidly growing industries, providing a stable supply of raw materials and market demand for the establishment of solar module recycling plants.
Europe
Many European countries were early adopters of photovoltaic (PV) technology and possess a substantial installed base; consequently, they are poised to gradually enter a peak phase for solar module decommissioning over the coming years. The European Union features a well-developed regulatory framework for waste management—including explicit requirements for e-waste recycling—which provides a solid policy foundation for the establishment of solar module recycling plant.
Countries such as Germany, Spain, and Italy—which rank among the leaders in PV installed capacity and possess mature supporting industrial ecosystems—are particularly well-suited locations for siting solar module recycling projects.
China
China stands as the world’s largest producer and installer of solar PV systems, boasting a cumulative installed capacity exceeding 500 GW. This signifies that a massive volume of end-of-life solar panels will be generated in the coming years, making the establishment of recycling facilities an effective means of processing discarded solar modules locally. Concurrently, relevant government authorities have recently begun actively promoting the development of standards and pilot programs specifically dedicated to solar panel recycling.
Selecting China as the location for solar module recycling plant allows for the full leveraging of local supply chain advantages, thereby ensuring the highly efficient operation of the recycling process.
United States
The U.S. solar market continues to grow, bolstered by strong support for renewable energy across various states. Some states have already begun to address the issue of end-of-life PV modules and are advancing relevant legislation. For instance, federal and state-level policies—such as the “Infrastructure Investment and Jobs Act”—provide financial support for recycling initiatives.
Establishing solar module recycling plant within the United States helps ensure compliance with domestic environmental regulations, offers manufacturers localized recycling services, and reduces transportation costs.
India
India’s installed solar capacity is climbing rapidly, having already surpassed 50 GW, with a government target of reaching 450 GW by 2030. This trajectory will generate a massive demand for recycling services.
Although India is an emerging market for photovoltaics—and its recycling infrastructure is currently still under development—proactively establishing solar panel recycling facilities will enable the country to meet future market demands while simultaneously fostering job creation and local technological advancement.
Australia
 
Australia boasts a high rate of solar energy adoption, alongside a steady increase in utility-scale ground-mounted solar projects. The government is actively promoting renewable energy recycling initiatives—such as the Solar Recycling Fund—which offer incentives for the construction of new recycling plants.
Setting up solar panel recycling plants in the vicinity of Australia‘s major cities or industrial zones would enable coverage of both local markets and select markets within the Pacific region.
These countries hold significant potential for solar panels recycling, as their market scale and policy environments are conducive to supporting the long-term operation of solar panel recycling plant. Naturally, specific investment decisions should be made in conjunction with local regulations and economic assessments. If you are considering investing in solar panel recycling plant in the aforementioned countries or other regions, we invite you to consult with the professional manufacturer of solar PV panel recycling machine: Henan DOING Company.
DOING Company’s solar panel recycling plants offer complete recycling production lines, encompassing dismantling, crushing, and sorting equipment. Utilizing the automated crushing and sorting system, our technology separates high-purity valuable materials—such as glass, silicon, aluminum, and copper—from discarded solar modules. With an extraction rate exceeding 98%, our solutions help recycling facilities minimize operational costs while ensuring compliance with international environmental standards.
Whether you plan to establish solar module recycling plant in Germany, Spain, Italy, China, the United States, India, or any other country, our equipment can be tailored to meet local requirements, and we are ready to provide customized recycling solutions to suit your specific needs.
Environmental XPRT is a global environmental industry marketplace and information resource. Online product catalogs, news, articles, events, publications & more.
Channels
Solutions
Latest
Publications
Companies

source

JinkoSolar has formally terminated its 4 GW Hai Ha solar cell project in Vietnam, citing US antidumping duties that undermined its export economics. The company will maintain other Vietnamese operations while shifting focus to localized supply chains and Asia-Pacific and Middle East markets.
Image: JinkoSolar
JinkoSolar’s Hai Ha solar cell project in Vietnam was formally terminated on March 11, after the investor requested to halt development and the Quang Ninh Provincial Economic Zone Management Board approved the request under Vietnam’s Investment Law. The authority also revoked the project’s investment certificate, No. 3282186668.
The project had been scheduled for development by Jinko Energy (Vietnam) Intelligent Manufacturing Co., Ltd., a vehicle set up by one of JinkoSolar’s Hong Kong subsidiaries, at Hai Ha Industrial Park in Quang Ninh province. The facility was planned to reach 4 GW of solar cell capacity and 3 GW of module capacity. The project initially carried a total planned investment of $1.5 billion when filed in October 2023, later revised to around $294.2 million in January 2024.
The termination followed a formal notice submitted by the company to the Quang Ninh administrative service center on Feb. 26, 2026. Vietnamese official reporting noted that the shutdown was the investor’s decision, rather than a government-mandated closure.
A key factor behind the termination was the worsening economics of exporting Vietnam-made PV products to the United States. In June 2025, the US Department of Commerce issued antidumping duty orders on crystalline silicon PV cells from Vietnam, following affirmative final determinations by both Commerce and the US International Trade Commission. Under the amended final determination, Jinko Solar (Vietnam) Industries Co., Ltd. was assigned an estimated weighted-average dumping margin of 125.91% and an adjusted cash deposit rate of 120.38%.
The U.S. trade measures appear to have undermined the strategic rationale for the Hai Ha project, which had been part of JinkoSolar’s Southeast Asian manufacturing footprint aimed in part at the US market. The termination reflects not only a project-level decision but also a broader reassessment of Southeast Asian solar manufacturing economics amid rising US tariffs.
The move does not signal a full withdrawal from Vietnam. JinkoSolar continues to operate other manufacturing facilities in the country, including solar cell and wafer plants, with around 10 GW of integrated n-type capacity remaining online. The company is also reportedly pivoting its global strategy toward more localized supply chains and greater focus on markets such as the Middle East and the wider Asia-Pacific region.
JinkoSolar is now expected to complete the project’s termination process, including handling construction in progress, tax and social insurance obligations, and land lease matters. The case highlights how trade policy is reshaping overseas manufacturing for Chinese solar companies, pushing investment decisions toward profitability, market access, and regional diversification rather than simple capacity expansion.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Vincent Shaw
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

Account Subscription: ACTIVE
Account Subscription: INACTIVE
Account Subscription: REGISTERED
Questions about your account? Our customer service team can be reached at [email protected] during business hours at (207) 791-6000.
Loading…
You are able to gift 5 more articles this month.
Anyone can access the link you share with no account required. Learn more.
With a The Portland Press Herald subscription, you can gift 5 articles each month.
It looks like you do not have any active subscriptions. To get one, go to the subscriptions page.
With a The Portland Press Herald subscription, you can gift 5 articles each month.
Mainers may soon be able to install small-scale, portable solar energy systems in their homes, under a bill backed by the Legislature on Thursday.
If signed into law, the measure, LD 1730, would allow electricity customers to use certain small solar generation and battery systems, which plug directly into wall sockets, similar to gas generators. Since they attach to a home’s electrical system like any other appliance, these panels are also portable, meaning homeowners and renters can take them along when moving — unlike traditional solar systems, which are generally permanent.
The proposal comes as Mainers face steep electricity prices, driven largely by the cost of natural gas. It also comes as the American and Israeli conflict with Iran shakes global oil markets.
Proponents argue that these small generators could offset a household’s electricity usage and lower their monthly utility bills, all for a significantly lower upfront cost than larger-scale, more traditional solar systems.
Lawmakers have offered different estimations on exactly how much one of these systems could save a typical Maine household, but it could be hundreds of dollars per year.
Rep. Gerry Runte, D-York, said last month that an 800-watt system could reduce an average Central Maine Power Co. customer more than 750 kilowatt hours each year, reducing their annual electricity costs by more than $250.
The Senate approved the measure without debate Thursday, sending it to Gov. Janet Mills.
The bill’s movement comes as plug-in solar systems see expanded interest and implementation globally. Twenty-eight states — including nearly all of New England — are currently considering similar bills, according to an analysis by Canary Media.
The United Kingdom announced last month that plug-in solar panels would “be in shops within months,” noting that they have already been adopted elsewhere in Europe. In Germany, more than one million panels have been installed since 2022, the local outlet DW reported late last year.
The new measure is expected to create a “minor cost increase” for the Public Utilities Commission, which could likely be absorbed by the commission’s existing budget, according to the bill’s fiscal note.
The measure would permit electricity customers to install systems with outputs up to 420 watts by themselves.
Larger systems, up to 1,200 watts, are also permitted, but they must be installed by a licensed electrician and attached to a dedicated circuit. Though they do not need to seek permission, residents who install larger systems are also required to notify their electric utility within 30 days of installation.
The bill prohibits electrical utilities from demanding applications, charging fees or requiring additional equipment, and it exempts utilities from any liability if the new systems cause damage or injuries.
None of these smaller systems can be used to earn net energy billing credits, according to the bill’s latest language.
In remarks ahead of the bill’s first vote last month, Rep. Gary Friedmann, D-Bar Harbor, said the bill “promotes the frugal self sufficiency that has kept us strong.”
But opponents in the Legislature have derided the measure as overregulation and argued that it could create safety concerns and lead to conflicts between renters and their landlords.
In a debate last month on the House floor, Rep. Reagan Paul, R-Winterport, said the bill “erodes property rights” by allowing renters to make changes to their living spaces without the property owner’s consent. She added that Maine’s old housing stock could create additional safety issues if, for example, someone runs an extension cord out their window because their home does not have exterior outlets.
Rep. Sophie Warren, D-Scarborough, argued that Mainers can already purchase appliances that would have a greater effect on a home’s electrical system than these solar and battery systems.
“This is about equity in our clean energy transition,” Warren said.
Staff Writer Rachel Ohm contributed to this report.
Daniel Kool is the Portland Press Herald's cost of living reporter, covering wages, bills and the infrastructure that drives them — from roads, to the state's electric grid to the global supply chains… More by Daniel Kool
We invite you to add your comments. We encourage a thoughtful exchange of ideas and information on this website. By joining the conversation, you are agreeing to our commenting policy and terms of use. More information is found on our FAQs. You can update your screen name on the member’s center.
Comments are managed by our staff during regular business hours Monday through Friday as well as limited hours on Saturday and Sunday. Comments held for moderation outside of those hours may take longer to approve.

Please sign into your Press Herald account to participate in conversations below. If you do not have an account, you can register or subscribe. Questions? Please see our FAQs.

source

source

India-based EV battery manufacturer Neuron Energy plans to enter the grid-scale storage market with a 5 GWh battery energy storage system (BESS) manufacturing facility in the Indian state of Maharashtra.
Image: Neuron Energy
From pv magazine India
Neuron Energy has announced plans to build a fully automated battery energy storage system (BESS) manufacturing facility in Talegaon, in the western Indian state of Maharashtra, marking its entry into the grid-scale storage segment.
The 7-acre facility is being developed with an investment of INR 1 billion ($10.8 million) and is designed as a robotic manufacturing plant for containerized energy storage systems. Once fully operational, it will have an annual production capacity of 5 GWh and the capability to produce up to 1,000 BESS units per year.
The modular systems are intended for deployment across solar and grid infrastructure, enabling storage of surplus energy during peak generation periods and dispatch during periods of high demand.
The company said the facility will create more than 500 direct and indirect jobs across engineering, manufacturing, system integration, installation, and technical services.
Neuron Energy said its go-to-market strategy will target solar developers, commercial and industrial (C&I) customers, and utilities, with a planned business mix of 60% domestic and 40% export markets.
The company added that it plans to expand its energy storage portfolio as it builds on its existing battery manufacturing capabilities.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Uma Gupta
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

Renewables equipment distributor Segen has launched a solar panel recycling scheme for commercial projects in the UK.
April 1, 2026
Renewables equipment distributor Segen has launched a solar panel recycling scheme for commercial projects in the UK. 
The company announced the plan today, to serve decommissioning and repowering needs for commercial solar installations. It said that as the UK’s solar capacity continues to grow, managing modules at the end of their life or when replaced with newer technologies is of increasing importance. 
Segen said it will recover silver, silicon and high-purity glass from decommissioned modules, which it said would then be “fed back into the manufacturing chain.” The company did not give further details of this claim, regarding where or how the materials would be returned to the PV manufacturing chain. 
The company will charge £6 for each panel below 2m in size and £7 for anything larger. It said it plans to launch a service for residential module recycling in “phase 2” of the scheme, which is currently under development. 
Related:SSE puts first solar PV plant online in England
Per media reports, Darren Sykes, head of warehouse & logistics UKI at Segen, said: “With the UK’s commercial solar capacity continuing to expand rapidly, more installations are reaching the end of their operational life, or being repowered as technology advances.
“As the UK’s largest renewables distributor, we consider it our responsibility to deliver a practical, reliable and cost-effective solution. Our recycling scheme will not only ensure panels are handled safely, but will also recover valuable materials for reuse, reduce landfill impact, and strengthen our customers’ ESG credentials.” 
Currently, solar modules in the UK are required to be recycled under the Waste Electrical and Electronic Equipment (WEEE) regulations, with some firms already offering recycling services. However, according to law firm Osborne Clarke, writing for Solar Power Portal last year, the cost of these recycling obligations can be complex, especially when considering modules which may have become defective and failed to reach their expected usable lifespan. 
PV recycling and end-of-life have become a major concern in many developed markets. Our sister site, PV Tech, has covered the matter extensively, most recently in a deep dive on the US module recycling industry (subscription required). Other markets, notably Australia, have also made significant efforts to deal with the growing amount of PV waste as the industry proliferates. 
Will Norman
Senior Reporter, Informa
Will is a senior reporter who primarily covers the policy and geopolitics behind the energy transition, with a particular focus on manufacturing. 
Featured Event
The US solar industry is evolving fast. Developers, investors, and utilities are navigating hybrid assets, multi-state pipelines, storage co-location, and complex grid challenges. SolarPLUS USA brings together the people, insights, and connections you need to scale reliable, clean power across the US.
Copyright © 2026 All rights reserved. Informa Markets, a trading division of Informa PLC.

source

Available in 80 W and 108 W versions using high-efficiency SunPower Maxeon cells, the kit offers wind resistance up to 25 m/s and easy installation on unused boat surfaces, according to the manufacturer.
Image: Solbian
Italian PV module manufacturer Solbian has unveiled a new solar kit designed for installation on boat davits.
“The solution provides both recreational sailors and marine professionals with a simple and efficient way to generate renewable energy on board, making use of surfaces that often go unused,” a company spokesperson told pv magazine.
The SunBoard kit features Solbian’s first rigid-frame module, engineered to provide a stable, deformation-free surface for davit installation. The panel is securely fastened to the upper davit and stanchions using ropes attached to perimeter eyelets, ensuring maximum stability even during storage or in strong gusts of wind, according to the company. The lower edge of the module is mounted on an aluminum rod with sliding clips, allowing the panel to be positioned and tilted between 0° and 90° for optimal sun exposure.
The solar module is available in two variants, SunBoard Kit S and SunBoard Kit M, both using high-efficiency SunPower Maxeon back-contact cells. Kit S delivers 80 W of power with a conversion efficiency of 24.0%, measuring 670 mm × 535 mm × 10 mm and weighing 2.2 kg. The larger Kit M produces 108 W at 24.4 % efficiency, sharing the same footprint as Kit S but slightly heavier at 2.8 kg.
Electrical specifications include maximum power voltages of 14 V (Kit S) and 19 V (Kit M), with currents at maximum power of 5.7 A. Open-circuit voltages are 16.7 V and 22.5 V, while short-circuit currents are 6 A. The panels operate in temperatures ranging from -40 C to 85 C, with a temperature coefficient of -0.27 %/C.
Each panel features 5 mm eyelets for secure attachment and dual junction boxes compatible with MC4 connectors, available in both port and starboard versions. Optional accessories include extension cables and deck penetrations. The system supports a maximum voltage of 1,000 V and a reverse current rating of 12 A, with a five-year manufacturer warranty.
According to Solbian, the kit is also engineered to withstand wind speeds of up to 25 m/s, making it a robust and flexible solution for on-board solar power generation.
 
 
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Emiliano Bellini
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

ALLPOWERS SF100 100W Flexible Solar Panel – IP68 Waterproof ETFE Solar Module For RV, Boat, Van, Roof  ruhrkanal.news
source

Japanese solar cell and module manufacturer Toyo shipped 4.5GW of cells in FY2025, surpassing its full-year target, while module shipments reached 249MW. 
The manufacturer reported revenue of US$427.4 million in FY2025, up 142% year-on-year, driven primarily by a US$241.6 million increase in solar cell sales and a US$7.6 million rise in module revenues. 
Get Premium Subscription
Toyo had set a solar cell shipment target of 4.2GW-4.4 GW. According to the firm, the exceeded target was driven by the ramp-up of its 4GW manufacturing facility in Ethiopia, which reached nameplate capacity in October 2025. 
The company also reported an increase in gross profit to US$96.3 million, with margins expanding to 22.5%, compared with US$21.9 million in 2024. Meanwhile, the net income came in at US$37.2 million, slightly down from US$40.5 million in the previous year. 
Toyo’s CEO Takahiko Onozuka said: “Our record revenues were underpinned by the rapid ramp-up of our 4GW cell facility, which is now operating at full capacity to serve our US utility-scale partners with high-efficiency, policy-compliant solar technology.” 
“Our production at our Houston module facility is expected to scale fast over the course of 2026 and we are evaluating additional strategic initiatives to create a robust onshore supply chain for US customers using advanced technology and performance standards,” he added. 
For 2026, Toyo has set shipment targets of 5.5GW-5.8 GW for solar cells and 1GW-1.3 GW for modules. The company forecasts adjusted net income of US$90-US$100 million for the full year. 
The CEO said Toyo will focus on establishing itself as a supplier of compliant solar solutions aligned with evolving US customer requirements. He added that solar remains a highly viable option for rapidly scaling energy production in a cost-effective manner.  
He also noted that the company had secured new polysilicon sourcing relationships and was advancing plans to onshore cell manufacturing in the US to support customers seeking integrated domestic content. Recently, Toyo secured a one-year supply agreement with an undisclosed firm for US-made polysilicon. 
As trade and tax-credit policy in the US tightens around the provenance of solar hardware, with non-domestically produced content facing closer scrutiny, Toyo said it was benefitting from having a transparent supply chain.
Toyo’s chief strategy officer, Rhone Resch, said: “Our outperformance this year is a direct result of our ability to navigate a complex global trade landscape. TOYO has built a resilient, traceable supply chain that the market trusts. As solar continues to drive the majority of new US electricity demand, TOYO is now well positioned with the domestic capacity and policy expertise to contribute to the next phase of the energy transition.”

source

© 2025 THE COOL DOWN COMPANY. All Rights Reserved. Do not sell or share my personal information. Reach us at hello@thecooldown.com.
If upcoming projects are realized, solar power would make up 51% of the new generating capacity.
Photo Credit: iStock
Although it may seem like clean energy projects in the United States have slowed, recently released data predicts 93% of the new utility-scale electricity generating capacity installed this year will come from solar, batteries, and wind power. 
The report from the U.S. Energy Information Administration, which tracks U.S. power plant developers and operators, revealed that there are plans to add 86 gigawatts of generating capacity to the grid in 2026 . 
If these projects are realized, solar power would make up 51% of the new generating capacity with battery storage following with 28% and wind at 14%. 
The data reveals 2026 will be a “big year” for grid-scale solar projects. 
Want to go solar but not sure who to trust? EnergySage has your back with free and transparent quotes from fully vetted providers in your area.
To get started, just answer a few questions about your home — no phone number required. Within a day or two, EnergySage will email you the best options for your needs, and their expert advisers can help you compare quotes and pick a winner.
“Developers plan to add 43.4 GW of new utility-scale solar capacity in 2026, a 60% increase in capacity additions from last year if realized,” it stated. 
These massive solar farms can help stabilize grids and reduce energy costs across communities, but smaller-scale residential solar is still the best way to reduce your individual energy bills while taking control of your power generation. 
If you’re interested in upgrading, check out EnergySage’s free tools to get quick installation estimates and compare competitive quotes. 
According to the numbers, more than half of the new large-scale solar projects are planned for just four states: Texas, Arizona, California and Michigan. The largest single solar project expected to power up in 2026 is in Texas, which will deliver 837 megawatts of power when completed. 
FROM OUR PARTNER
Want to go solar but not sure who to trust? EnergySage has your back with free and transparent quotes from fully vetted providers that can help you save as much as $10k on installation.
To get started, just answer a few questions about your home — no phone number required. Within a day or two, EnergySage will email you the best local options for your needs, and their expert advisers can help you compare quotes and pick a winner.
Although this data suggests energy developers are taking steps to transform the U.S. grid into a clean energy powerhouse, you don’t have to wait for a large-scale project to come to your area to benefit from solar panels. 
In fact, residential solar is one of the most reliable bill-saving investments you can make. Some homeowners can even achieve six figures in savings over the lifetime of their solar system. 
Which of these savings plans for rooftop solar panels would be most appealing for you?
Save $1,000 this year 💸
Save less this year but $20k in 10 years 💰
Save less in 10 years but $80k in 20 years 🤑
Couldn’t pay me to go solar 😒
Click your choice to see results and earn rewards to spend on home upgrades.

To fully understand how much getting solar can cut down your energy bills, check out TCD partner EnergySage. With its free tools, you could save up to $10,000 on solar purchases and installation costs. 
EnergySage even offers a mapping tool that shows you all of the incentives available in your area and the average costs for solar panels on a state-by-state level. With it, you can find the best price for rooftop panels while ensuring you lock in all of the rebates and tax credits you can get. 
💡Go deep on the latest news and trends shaping the residential solar landscape
While you’re considering solar panels, you may want to think about making your home energy even more secure by installing a battery backup. Home energy storage is one of the most effective ways to protect your home from outages, save even more on energy bills and, if you want, completely cut ties with your local grid. 
To learn more about pairing batteries with solar, and get competitive installation quotes, check out EnergySage’s free storage resources. 
Get TCD’s free newsletters for easy tips to save more, waste less, and make smarter choices — and earn up to $5,000 toward clean upgrades in TCD’s exclusive Rewards Club.
© 2025 THE COOL DOWN COMPANY. All Rights Reserved. Do not sell or share my personal information. Reach us at hello@thecooldown.com.

source

DENVER, CO – Legislation to remove barriers to plug-in solar panels and save Coloradans money on their utility bills passed the Senate today. 

HB26-1007, sponsored by Senators Cathy Kipp, D-Fort Collins, and Matt Ball, D-Denver, would authorize access to plug-in solar panels, which can be plugged into a home electrical outlet and are more affordable than traditional rooftop solar.

“This bill reduces barriers and establishes safety standards so that Coloradans who want a reliable, affordable source of renewable energy can use plug-in solar panels,” said Kipp. “Coloradans are interested in plug-in solar for a variety of reasons like reducing their carbon footprint, lowering their utility bills, or ensuring a reliable back-up source of energy in the case of a power outage. No matter their reasoning, Coloradans should be able to pursue this technology without unnecessary barriers.” 

“Plug-in solar panels expand access to solar energy for people who live in an apartment or can’t afford a full rooftop system,” said Ball. “The technology is safe, cost-efficient, and already widely used in other places. This bill gives Coloradans the option to use plug-in solar and connect to the grid through a meter collar to start saving money and producing their own clean energy.” 

Plug-in solar, also referred to as balcony solar, can be plugged into a home electrical outlet and is more affordable than traditional rooftop solar. It consists of one to four solar panels plus an inverter and optional battery and is designed for simple, safe installation. Plug-in solar can be used to power household appliances and offer Coloradans an alternative, reliable energy source that can also reduce traditional utility costs.

The bill would establish protective guardrails on the types of plug-in solar products that can be used. Under this bill, all plug-in solar devices installed must meet the UL 3700 product safety standard. 

HB26-1007 would also encourage the use of meter collars. Meter collars are devices installed between an electric meter socket and a utility billing meter to provide immediate interconnection of customer-owned solar devices to the grid. Meter collars eliminate the need for a costly electrical panel upgrade, saving Coloradans money and time on solar installation. This bill outlines a safe, consistent and repeatable solar installation process with minimal disruption and short installation times to benefit Coloradans. 

Plug-in solar is common in Europe. For example, in Germany, approximately 4 million households have installed plug-in solar. If passed, Colorado would join Utah in becoming early adopters of safe, reliable, plug-in solar in the United States. 

HB26-1007 now heads back to the House for consideration of amendments. Track its progress HERE.
Senators
Priorities
Newsroom
Job Opportunities
2025 Session Achievements
Colorado General Assembly

source

Xinhua | January 12, 2026
China announced on Friday that it will change export tax rebates for a range of products, including photovoltaic and battery products.
The announcement, jointly issued by the Ministry of Finance and the State Taxation Administration, said that export tax rebates for the value added tax of photovoltaic products will be canceled starting from April 1, 2026.
Meanwhile, the export tax rebate rate for the value added tax of battery products will be reduced from 9 percent to 6 percent starting from April 1, 2026, and will be eliminated starting from Jan. 1, 2027, according to the announcement.
The move is welcomed by China's domestic industry, with the China Photovoltaic Industry Association saying that the measure will help restore rational pricing in foreign markets and reduce the risk of trade frictions for China.
Over the long term, the measure will help prevent export prices from falling too rapidly and lower the risk of trade disputes further, according to the association. 

source

Published daily by the Lowy Institute
US tariffs are testing whether ASEAN is all talk on regional integration for renewable energy generation.
Southeast Asia’s solar panel manufacturers are facing hard times. Having benefited from anti-China trade measures, Southeast Asia had become the largest source of US solar photovoltaic (PV) imports – until Washington turned on the region. Once bright solar ambitions are beginning to set. Southeast Asia needs to find answers closer to home.
The United States has persistently raised trade barriers to solar panels for more than a decade but the key factor driving import flows has been the tariff differential between countries. As president, Barack Obama imposed solar panel duties on China, and Donald Trump imposed general goods tariffs during his first term.
This gave Chinese firms an incentive to set up manufacturing operations in Southeast Asia. Vietnam, Thailand, and more recently Cambodia were the prime beneficiaries. This added to significant solar PV manufacturing in Malayia. Solar exports to the United States surged from US$2.8 billion in 2012 to US$14.4 billion by 2023. By 2024, 80% of US solar PV imports came from these four countries.
However, dependency cuts both ways. The United States represented the same share of Southeast Asia’s total solar exports, leaving the industry vulnerable.
ASEAN has trumpeted deepening regional economic integration as a response to tariffs. Solar would be a good place to start.
Viewed as an extension of Chinese overcapacity, Joe Biden effectively ended Southeast Asia’s access to the US market in 2024 by imposing anti-dumping and countervailing duties. Trump announced even higher import duties in April of this year. Now the average effective rate for Malaysia is roughly 34%, while the determination on Vietnam and Thailand is well over 300%, and Cambodia is hit hardest with 652%. Firm specific as well as general rates for each country’s industry were imposed, meaning rates reached as high as 3,400% for some firms. These will translate into countervailing duties (effectively tariffs) alongside additional anti-dumping duties ranging from 2% to 271%.
The industry has already started showing signs of stress. More job losses and factory closures are expected. Even Malaysia’s solar manufacturers are struggling despite receiving the lowest tariff rates.
Much of the region’s manufacturing capacity is underpinned by Chinese investment, inputs, and technology. China accounted for more than 80% of foreign investment into new projects. This reflects China’s dominance in global solar PV and global supply chains.
But for the countries in Southeast Asia, China’s omnipotence in their solar PV industry is exaggerated. The United States Uyghur Forced Labour Prevention Act and other measures were used to prevent polysilicon (the upstream material needed to manufacture solar cells) from China’s Xinjiang region being used in US-deployed solar due to human rights concerns. Malaysia developed and expanded domestic polysilicon production while Vietnam began importing polysilicon from Germany, the United States, and Malaysia.
In response to Biden-era tariffs, Indonesia and Laos became the next best locations to export from. Indonesia brought online 20 gigawatts of foreign-owned solar manufacturing capacity, up from just 1 gigawatt at the end of 2022. China led the way with manufacturing investments in these countries, to be sure. But US investments in Indonesia since 2023 are about one-third the value as those from China, and in Laos, the figure is approximately half. This shows Southeast Asia’s solar developments are not solely dictated by Beijing.
Unfortunately, the Trump administration announced new anti-dumping and countervailing duties investigations into Indonesia and Laos in August, the outcomes of which are easy to guess. Washington is also targeting upstream supply chain segments with a national security trade investigation into polysilicon and derivative products. Any hopes that Southeast Asia might continue exporting to the US market are rapidly fading.
Trump’s trashing of US domestic green energy policies will inevitably slow down renewable energy capacity additions, so export diversification was always necessary for Southeast Asia’s manufacturers. But now it is urgent. Competing in China is not an option. The two biggest markets outside China and the United States are Europe and India. Europe, the largest solar importers, almost exclusively import from China. India emulated China with a Make in India policy, which included a domestically grown solar PV industry that quickly developed 100GW of solar panel manufacturing capacity while doubling down on protectionist interventions. Meanwhile, China’s export volumes continue surging.
Southeast Asia will have to rescue itself.
Renewable energy generation globally will double over the next five years and 80% will be solar PV. Accelerating the region’s energy transition will support local manufacturers. Southeast Asia is already the fourth largest source of global energy demand, will drive one-quarter of demand growth to 2035, and holds the largest solar PV module production capacity outside of China at 106 GW, with expanding production across other supply chain segments. Under already announced policies, the region wants to install 28 GW of solar PV and bring total capacity to 100 GW by 2030. Yet these investments in utility-scale solar are not even keeping pace with other major emerging economies such as India and Brazil.
If Southeast Asia adopted a net-zero policy pathway, the International Energy Agency estimates it would need more than 250 GW. Reaching this level would create demand, protect jobs and factories, and generate new employment as rolling out the panels inevitably employs more people than solar PV manufacturing.
Regional coordination would obviously scale such actions. The ASEAN Power Grid initiative already provides an institutional mechanism to accelerate solar deployment for the entire region through electricity grid integration.
ASEAN has trumpeted deepening regional economic integration as a response to tariffs. Solar would be a good place to start.
Get the latest commentary and analysis on international events from experts at the Lowy Institute and around the world.
The Interpreter features in-depth analysis & expert commentary on the latest international events, published daily by the Lowy Institute.
© Copyright 2026 Lowy Institute

source

Platts extends solar PV price assessments in India  spglobal.com
source

Guidelines to encourage the comprehensive utilization of photovoltaic (PV) modules have jointly been released by six Chinese national authorities. These include Ministry of Industry and Information Technology, Ministry of Ecology and Environment, Ministry of Commerce, State Administration for Market Regulation, National Financial Regulatory Administration, and National Energy Administration.
Key tasks of the guidelines include refining relevant laws, regulations, policies and standards, boosting R&D of processing technologies, broadening application channels for related products, and strengthening support for key production factors. These will help foster organized development of China's comprehensive PV module utilization sector.
By 2027, the country will further raise green production levels of PV modules, effectively increase the proportion of recycled materials in production, and refine the evaluation criteria and inspection methods of module retirement, according to the guidelines.
In addition, breakthroughs in core technologies including surface structure disassembly, efficient separation of PV laminates, and component extraction will be achieved. The comprehensive utilization of retired PV modules will be further scaled up in key sectors including metal smelting, equipment manufacturing and building materials production.
A series of technical standards for the green design and comprehensive utilization of PV modules will be developed, a batch of leading enterprises in the sector will be fostered, and the cumulative volume of PV modules processed through comprehensive utilization will reach 250,000 tonnes.
By 2030, the technological and equipment level for comprehensive utilization of PV modules will be further enhanced with significantly strengthened industrial innovation and development capabilities, the guidelines noted. Application scenarios and methods for comprehensive utilization products will continue to expand, creating a strong capacity for using end-of-life PV modules. This capacity will feature close coordination across the industrial chain, rational production capacity layout, and the ability to address large-scale retirement waves.
Efforts should be made to advance the green design and manufacturing of the PV industry, improve the ease of dismantling and recycling of PV modules, and raise the proportion of recycled materials used in production, according to the guidelines. It calls for the orderly decommissioning of end-of-life PV modules, and guides relevant stakeholders to standardize the handover and delivery of waste PV modules.
The guidelines also stipulate the promotion of green and high-efficiency dismantling and utilization, encourage the development of non-destructive dismantling technologies, and support the extraction of silver materials from the metal grid lines of crystalline silicon solar cells.
Furthermore, the guidelines advocate promoting the coordinated development of the entire industrial chain for the comprehensive utilization of PV modules, and encourage PV module manufacturers, solar power station operators, and comprehensive utilization enterprises to actively extend their industrial chains.
Efforts will be made to optimize the environment for industrial innovation and development. Management measures for comprehensive utilization of industrial resources will be accelerated to clarify the responsibilities of all parties involved in the comprehensive utilization of waste photovoltaic modules. Policy support will also be strengthened.
Qualified PV industrial parks are encouraged to develop "waste-free zones." Support will therefore be given for collaboration among PV module manufacturers, comprehensive utilization product producers, and end-users to develop exemplary green and low-carbon industrial practices within the PV sector.
Enterprises in the PV module comprehensive utilization industrial chain will be supported to strengthen cooperation with internationally advanced companies. This will involve exchanges and learning from each other in areas such as technology, talent, and management model innovation. Ultimately, it would help improve their international competitiveness.
Chinese President Xi Jinping has sent a congratulatory letter to the Science and Technology Daily on the occasion of the 40th anniversary of its founding.
​On March 30, 2026, the Bayuegua Institute of Science and Technology Innovation, a sci-tech innovation think tank service institution, officially released the Global AI Enterprise Technology Innovation Index Report 2026.
6G Development in Critical Phase
High-Altitude Sentinel in Global Climate Fight
Sci-tech Central to Consumption Upgrades
Robots Handle Tough Workplaces With Ease
Mapping the Sky: Dunhuang Star Chart
Development of 4th Gen OLED Screens

source

UAE’s global projects pave the way for a clean energy future  Gulf Today
source

The Dutch research institute has presented what it describes as the world’s first perovskite-based roof tile, achieving up to 13.8% efficiency on standalone modules and 12.4% when installed on a curved surface. The flexible modules were produced using TNO’s experimental roll-to-roll platform,
Image: TNO
The Netherlands Organization for Applied Scientific Research (TNO) has unveiled today a building-integrated photovoltaic (BIPV) tile based on perovskite solar cell technology.
The new product is billed as the world’s first perovskite solar tile.
“This demonstrator is supported by the Province of North Brabant through the project ‘Solar manufacturing industry to Brabant, Solliance 2.0’. Additional funding was received from the European Union’s Horizon Europe programme for the Luminosity project,” TNO said in a statement. “The work was also partly funded by the National Growth Fund programme SolarNL.”
The Dutch research institute partnered with Netherlands-based BIPV specialist Asat BV in deploying 10 cm x 10 cm perovskite solar modules built on flexible foil onto a curved composite roof tile. Testing indicates that bending the modules to fit the curved surface has minimal impact on their performance.
Standalone modules reached energy conversion efficiencies of up to 13.8%, while the modules retained an efficiency of 12.4% after installation on the curved roof tile.
Image: TNO
The perovksite modules were encapsulated with an experimental roll-to-roll manufacturing platform developed by TNO itself. Roll-to-roll manufacturing – similar to the process used in newspaper printing – enables continuous production of solar cells on long rolls of flexible material. The technique is widely seen as a potential pathway to lower production costs and high-volume manufacturing for emerging thin-film technologies such as perovskites.
More technical details about the solar tile were not disclosed. TNO said it will be commercialized by its spinoff Perovion Technologies, which was launched last month. 
TNO’s recent research on perovskite solar cells, includes developing roll-to-roll and spatial atomic layer deposition (SALD) processes for the deposition of functional materials, solar cell layers, and flexible foils.
In July, Solarge, a manufacturer of lightweight silicon PV modules based in the Netherlands, and TNO unveiled a 32 cm x 34 cm lightweight prototype perovskite solar panel.
A month earlier, Japan’s Sekisui Solar Film, part of Sekisui Chemical, the Brabant Development Agency (BOM), which serves the Dutch province of Noord-Brabant, and TNO signed a letter of intent in Osaka, Japan to explore collaboration related to flexible perovskite solar PV module technologies.
As pv magazine has reported, Sekisui Solar Film is developing technology for lightweight, flexible perovskite solar module manufacturing using an advanced roll-to-roll process. It is working on a 100 MW plant in Japan for large-scale production, is undertaking field demonstrations, and signed a perovskite solar-related memorandum of understanding with Slovakia.
 
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Emiliano Bellini
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

Distributed solar additions have vastly outpaced all other forms of generation as Puerto Rico’s overall power generation capacity continues to grow.
Image: US EIA
From pv magazine USA
Recently released data from the US Energy Information Administration (EIA) indicates that 20% of all power generation capacity in Puerto Rico now comes from rooftop solar, surpassing natural gas to become the second-largest capacity source in the territory.
Growth in rooftop solar capacity has outpaced all other energy sources in Puerto Rico over the past decade. According to EIA data, distributed solar installations accounted for 81% of all new generating capacity added to the island’s grid between 2016 and 2025.
During 2025 alone, an average of 3,850 rooftop systems were installed each month at homes and businesses, bringing the total number of active systems to 191,929 by year-end.
Rooftop solar capacity growth has been a bright spot in Puerto Rico’s energy story. The 1,456 MW of rooftop capacity far exceeds the estimated 165 MW of utility-scale solar installed on the island.
PJ Wilson, president of the Solar Energy and Storage Association Puerto Rico, said the industry group remains committed to expanding distributed solar across the territory.
“We are committed to building on this momentum and ensuring rooftop solar and storage continue to grow as a key part of Puerto Rico’s energy system to strengthen the grid and expand energy independence,” he told pv magazine USA.
Notably, the growth in solar capacity has not displaced other generation sources, with capacity from petroleum, natural gas, and coal showing little change over the past five years.
In 2025, Puerto Rico Governor Jenniffer González Colón signed Act 1-2025 into law, extending the lifespan of the territory’s only coal-fired power plant through 2032, despite opposition from local communities. The legislation also revised renewable portfolio standards, removing interim targets of 40% by 2025 and 60% by 2040, while retaining the long-term goal of 100% renewable energy by 2050.
Battery storage and virtual power plants
Grid resilience has become increasingly important in Puerto Rico in recent years. Data shows that the average utility customer experiences a minimum of 27 hours of outages annually, with some locations facing up to nearly 200 hours depending on severe weather events.
In response, adoption of distributed energy storage has grown rapidly. The Puerto Rico Energy Bureau estimates more than 171,000 households and businesses had installed battery systems by the end of 2025, representing a combined capacity of 2,864 MWh.
Analysts at Wood Mackenzie expect an additional 3,000 MWh of distributed storage to be added by 2030.
Many battery owners participate in virtual power plants (VPPs) through the Customer Battery Energy Sharing (CBES) program operated by grid operator LUMA. Through the program, LUMA works with storage aggregators that manage fleets of customer-sited batteries across the territory. During periods of peak demand, the utility can call on these aggregators to temporarily control distributed batteries in “CBES events” to help balance the grid.
LUMA currently lists seven aggregators on its website, allowing customers to enroll and receive compensation in exchange for participation.
Grid challenges have also prompted changes at LUMA. New CEO Janisse Quiñones began her tenure on March 30, 2026, bringing experience from her previous role as CEO and chief engineer of LADWP, with a stated focus on improving grid reliability.
Wilson said the industry group is optimistic about increased collaboration under new leadership.
“SESA remains focused on advancing policies that allow rooftop solar and battery storage to keep growing as a central pillar of Puerto Rico’s energy future, and we’re encouraged by the opportunity for stronger collaboration under LUMA’s new leadership,” he said.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Ben Zientara
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

Please enable JavaScript to proceed.

source

Industry Hub & Blog for B2B Industry – Mechanical Engineering – Logistics/Intralogistics – Photovoltaics (PV/Solar)
For Smart FACTORY | CITY | XR | METAVERSE | AI | DIGITIZATION | SOLAR | Industry Influencers (II) | Startups | Support/Consulting
Xpert Pre-Release
Language selection 📢

Published on: April 3, 2026 / Updated on: April 3, 2026 – Author: Konrad Wolfenstein
The end of cheap solar panels? Why modules from China are suddenly becoming more expensive – Image: Xpert.Digital
Why Beijing is now pulling the emergency brake – and the world is paying the price
For years, homeowners and stakeholders in Europe's energy transition benefited from an unprecedented drop in the price of photovoltaic systems. Solar panels and balcony power plants were cheaper than ever before – fueled by massive government subsidies and enormous overproduction in China. But this era of "cheap solar" is now coming to an abrupt end. With a far-reaching political U-turn, Beijing is radically cutting billions in export subsidies for its solar industry. What at first glance appears to be a distant fiscal policy measure has direct and tangible consequences for the domestic market: Prices for solar technology will rise noticeably. But why is China, the undisputed global market leader, pulling the emergency brake now? The following article examines the background of this historic decision – from the destructive price spiral of the so-called "Neijuan" to geopolitical power plays and the question of what this development means specifically for consumers' wallets and the future of the energy transition.
Related to this:
As of April 1, 2026, China has implemented a quiet revolution in the global solar market. The Chinese Ministry of Finance and the State Revenue Administration jointly announced the complete elimination of value-added tax (VAT) refunds for the export of photovoltaic products. This step affects all key components of the solar value chain – from monocrystalline silicon wafers and solar cells to fully assembled modules and inverters. A phased approach applies to energy storage products: The refund rate will initially be reduced from 9 to 6 percent and completely eliminated from January 1, 2027.
What at first glance appears to be a technical tax adjustment is in reality the culmination of a decades-long subsidy program. As early as December 2024, China had reduced the tax refund rate for photovoltaic products from 13 to 9 percent – ​​a first warning sign, now followed by its complete elimination. In barely three years, the state support framework for solar exports has thus been reduced from 13 percent to zero, unequivocally demonstrating the fundamental reorientation of Chinese industrial policy in this sector.
To understand the implications of this decision, one must grasp the extent of Chinese market dominance in the photovoltaic industry. China currently controls over 95 percent of global polysilicon production for solar applications, 97 percent of wafer manufacturing, 85 percent of solar cell production, and approximately 75 percent of module production. This near-complete control of all stages of the value chain is no accident, but rather the result of a two-decade-long industrial policy that combined government subsidies with massive capital inflows, favorable land prices, and coordinated technology promotion.
The production capacities resulting from this expansion are unprecedented. In 2025, China's manufacturing capacity for solar modules is estimated to have reached 1,200 gigawatts – a figure almost double the total global installation demand of around 650 gigawatts in the same year. In the first half of 2025 alone, China installed 212 gigawatts of new solar capacity, equivalent to the total photovoltaic capacity built in Germany over 25 years. By the end of 2025, China's cumulative photovoltaic capacity had surpassed the historic mark of 1,200 gigawatts – the first country in the world to do so.
Behind these impressive figures lies a structural crisis known in the People's Republic as Neijuan – a term from agricultural sociology that originally described stagnation despite increasing resource input, but today stands for destructive cutthroat competition without productive progress. In the solar industry, Neijuan has taken a concrete and mathematically verifiable form: manufacturers systematically sell below their cost to defend market share and finance the resulting losses with cheap government loans and provincial subsidies.
The consequences are dramatic. The four largest Chinese module manufacturers – Longi, Jinko Solar, Trina Solar, and JA Solar – recorded combined net losses of 11 billion yuan in the first half of 2025 alone, equivalent to approximately US$1.54 billion and representing a 150 percent increase compared to the same period of the previous year. Longi Green Energy, once the undisputed global market leader, reported losses of up to €700 million for the first half of 2025, while Tongwei warned of losses of up to €400 million in the same period. Jinko Solar recorded a 32.63 percent drop in revenue alongside exploding losses.
This price war developed its own dynamic: due to oversupply, solar modules became almost 50 percent cheaper over the course of 2024, which, while accelerating the global energy transition, drove Chinese manufacturers into an existential spiral. Despite a record expansion of 315 gigawatts of new solar capacity in China alone in 2025, the industry remained highly unprofitable – an economic paradox that exposes the limits of state-enforced overproduction.
The abolition of export subsidies is not a spontaneous reaction, but the result of a multi-year situation analysis at the highest governmental level. Several factors drove the decision at this time.
First, the industry's losses have reached a politically intolerable level. Beijing is no longer able to accept systematic losses in the billions as the price of global market dominance when its strategic goals—cost reduction, technological leadership, and market penetration—have already been achieved. Lin Boqiang, director of the China Center for Energy Economics Research at Xiamen University, describes this step as a necessary intervention to curb inefficient competition, reduce trade conflicts, and steer the industry toward high-quality development.
Secondly, China faces a growing problem with international trade disputes. The EU, the US, and an increasing number of other importing countries have initiated anti-dumping proceedings or imposed punitive tariffs, explicitly arguing that Chinese manufacturers are only able to export below market prices due to state subsidies. The China Photovoltaic Association (CPIA) confirmed in an official statement that the measure is intended to help rationalize pricing in foreign markets and reduce the risk of trade conflicts.
Third, the entire economy is trapped in a deflationary cycle, fueled by serial overproduction in several key industries—solar panels, batteries, and electric vehicles. Nomura analysts interpreted the decision as a signal that Beijing will rely more on non-currency instruments to manage its massive trade surplus problem, rather than resorting to yuan appreciation. The decisive end to subsidies also demonstrates a serious intention to enforce market correction in multiple sectors simultaneously.
 

New: Patent from the USA – Install solar parks up to 30% cheaper and 40% faster and easier – with explanatory videos! – Image: Xpert.Digital
The core of this technological advancement is the deliberate departure from conventional clamp mounting, which has been the standard for decades. The new, more time- and cost-effective mounting system addresses this with a fundamentally different, more intelligent concept. Instead of clamping the modules at specific points, they are inserted into a continuous, specially shaped support rail and held securely in place. This design ensures that all forces – whether static loads from snow or dynamic loads from wind – are distributed evenly across the entire length of the module frame.
More information here:
 
Alongside the elimination of export subsidies, Beijing is attempting to curb neiyuan competition through industrial policy coordination. In July 2025, the Ministry of Industry and Information Technology invited 14 major solar companies to a meeting where measures against irrational low-price competition, the decommissioning of outdated capacity, and improvements in industrial quality were agreed upon. The wording was the official terminology for what is internationally known as cartel coordination: coordinated production restrictions, minimum price agreements, and the coordinated closure of older manufacturing facilities.
This strategy carries significant domestic political tensions. Six leading polysilicon manufacturers, along with the industry association CPIA, were summoned by the state market regulator because their self-regulation agreements raised suspicions of illegal price-fixing. The authorities face a fundamental dilemma: effectively combating overcapacity requires coordination, which, without clear boundaries, risks veering into market manipulation. Industry experts anticipate that consolidation in silicon, wafer, and module manufacturing will take months or even years before its effects become noticeable.
Related to this:
The elimination of export subsidies is hitting an industry already suffering from increasing cost pressures. Silver, a key material in solar cell manufacturing, reached a record price of $83.62 per ounce at the end of 2025 – a year-on-year increase of over 130 percent. This means that silver paste now accounts for 15 to 17 percent of the total cost of a solar module, surpassing polysilicon as the largest single cost factor. According to OPIS analysts, FOB prices for TOPCon cells – the technologically leading product generation – had already risen by around 30 percent since mid-December 2025, even before the tax reform officially took effect.
In parallel, the Chinese Ministry of Industry has curbed polysilicon production through regulatory intervention. Leading manufacturers are now producing at only 55 to 70 percent of their capacity, which led to a 48 percent price increase for this key raw material in September 2025 alone. Wood Mackenzie had already predicted price increases of around 9 percent for modules and storage systems starting in the fourth quarter of 2025 – an assessment that has proven too conservative with the complete elimination of subsidies.
Since nearly 90 percent of solar modules sold in Germany are of Chinese origin, Beijing's tax policies are directly reflected in end-customer prices. Market analysts anticipate a price increase of 10 to 15 percent for solar modules in the European market. For mid-range systems, typically installed on single-family homes and currently costing between €15,000 and €18,000, this translates to an additional cost of around €600. In some sub-segments, price increases of 20 to 30 percent were already observed before the effective date, as European buyers attempted to protect themselves against the anticipated price hikes by placing pre-orders.
For balcony power plants – a booming entry point into decentralized energy generation in Germany – the elimination of Chinese export subsidies means a price increase of around 10 percent, according to industry representatives. A device currently available for 600 euros could rise to as much as 660 euros by 2026. Current export prices for TOPCon modules are already between 0.09 and 0.13 US dollars per watt-peak – and trending upwards.
For large-scale systems with 15–18 kWp, a price increase of approximately 3–4% is expected, corresponding to an additional cost of around €600. Standard solar modules could become 10–15% more expensive; the absolute additional costs depend on the respective power class. Balcony power plants, which currently cost around €600, are expected to see an increase of about 10%, i.e., additional costs of around €60. Module prices (FOB, TOPCon) have already risen by up to 30% since December 2025 and are currently around USD 0.09–0.13 per watt-peak. For battery storage systems, a reduction in the subsidy from 9% to 6% is expected in 2026, leading to a gradual price increase; in 2027, the subsidy will fall to 0%, resulting in a complete price increase for storage systems.
China's decision also follows a geopolitical calculation. In a world of increasing trade tensions, US tariff increases under President Trump, and growing EU skepticism towards Chinese subsidy practices, eliminating its own export tax breaks sends a rhetorically effective signal: Beijing is pretending to become more market-compliant and to address international accusations of dumping. At the same time, the Chinese solar industry has long since achieved its strategic goals – global market penetration, cost reduction through economies of scale, and technological leadership – and no longer needs subsidies to remain unrivaled.
The complete market dominance of Chinese manufacturers remains structurally intact even after the subsidies are eliminated. No Western manufacturer has the capacity to operate the value chain from polysilicon to the finished module competitively. Even a 15 percent price increase will only bring Chinese modules back from a historically unprecedented low to a still very affordable level – the cost advantage over alternative sources remains.
Behind the operational decision to eliminate tax breaks lies a deeper strategic realignment. The Chinese government has signaled its intention to transform the solar industry from a volume-oriented mass exporter into a technology-leading, high-value segment. Liu Yiyang, Executive Secretary General of the China Photovoltaic Association, described this transformation as follows: The end of tax privileges marks the point at which the industry must prove itself in free market competition. Companies that were only viable through government subsidies will exit the market; the remaining market leaders will emerge from this consolidation phase technologically stronger and more financially stable.
In this context, the shift towards more advanced cell technologies such as TOPCon, HJT, and perovskite tandem concepts is also understandable. Where cost reductions for conventional PERC cells reach their physical limits, these technologies offer new efficiency gains and thus the basis for a renewed price offensive at a technologically higher level. China installed a new world record of 315 gigawatts in 2025 and achieved its national expansion target of 1,200 gigawatts of installed solar and wind power six years ahead of schedule in 2030.
Eliminating subsidies will not automatically solve the Neijuan problem. As long as overcapacity exists in the value chain, the structural pressure for aggressive pricing will remain – only now without government funding to cover the shortfall. The market consolidation process will be painful: analysts expect significant capacity reductions, bankruptcies of medium-sized manufacturers, and government-orchestrated mergers to bring the industry to a sustainable size.
For international buyers and project developers, this results in a significantly altered basis for calculation in the medium term. The unusually low module prices of the years 2023 to 2025 were likely not a permanent state, but rather the temporary result of a flood of prices created by industrial policy. A return to these price levels is not realistic in the medium term – even if the absolute cost reduction of the technology through further efficiency improvements will enable further price reductions in the long term.
For the European and German energy transition, this is not a catastrophe, but a significant recalibration. The economic advantages of photovoltaic systems remain even with a price increase of 10 to 15 percent, as the levelized cost of electricity is still significantly lower than that of fossil fuel alternatives. What is changing is the pervasive logic of permanent price reductions, on which many installation companies, project developers, and end customers had based their financial plans. China's new solar economy is more expensive – and more honest.
 
 

Konrad Wolfenstein
I and my team are happy to be available to you as your personal advisor.
You can contact me by filling out the contact form here or simply call me at +49 7348 4088 965. My email address is : [email protected]
I'm looking forward to our joint project.
 
 
 

Innovative photovoltaic solution for cost reduction and time savings – Image: Xpert.Digital
More information here:
Your partner in Germany and Europe
 
 
© April 2026 Xpert.Digital / Xpert.Plus – Konrad Wolfenstein – Business Development

source

Solar photovoltaic’s are central to NASA’s Artemis missions, serving as the primary power source for the Orion spacecraft and future Gateway lunar outpost. Orion relies on four large solar array wings, which deploy in space to generate more than 11 kW. That’s enough to power life support, navigation, and communications.
The European Service Module powering Orion features four 7-meter-long solar arrays manufactured by Airbus. These arrays convert sunlight into electricity and maintain energy supply throughout the 10-day Artemis II mission, including the lunar fly-by. They rotate on two axes to track the Sun, ensuring continuous power and charging onboard batteries for periods when the spacecraft is in shadow. Each array contains 15,000 gallium arsenide photovoltaic cells, producing roughly 11.2 kW of power.
Future Artemis missions will use Roll-Out Solar Arrays on the Gateway outpost orbiting the Moon. These flexible arrays are designed to provide reliable power for long-term operations. NASA is also developing vertical deployable solar arrays for lunar surface exploration, optimising sunlight collection at the poles where the Sun remains low on the horizon.
During the Artemis I mission, Orion’s solar arrays has so far exceeded expectations, generating 15% more power than planned, demonstrating the efficiency and resilience of the technology.
Author: Bryan Groenendaal







Δ

February 26, 2026
March 18, 2026
February 20, 2026
March 16, 2026
February 25, 2026
March 16, 2026
Disclaimer | Privacy Policy | Terms & Conditions | Returns Policy | Intellectual Property | Cookie Policy
© 2019 – 2026 GBA Digital Media Group. All Rights Reserved | Site Credit
Copyright Green Building Africa 2024.

Subscribe to our weekly Top 5 Stories
"*" indicates required fields
Δ

source

Shared Video
Shared Video
Shared Video
Shared Video
Shared Video
Shared Video

source

Shared Video
Shared Video
Shared Video
Shared Video
Shared Video
Shared Video

source

The war goes on, and so does the global energy crisis. In fact, I believe that prices of oil futures remain too low given how much spot prices will need to rise to resolve the shortages that will hit once oil supplies that were shipped before the Strait of Hormuz was closed are exhausted.
But a better future is coming, despite Donald Trump’s assault on renewable energy as he tries to drag us back into the fossil fuel past. Regardless of Trump’s chest-thumping, America is not the world. We account for only 15 percent of global energy consumption, compared with China’s 28 percent. And the rest of the world is moving rapidly to renewables, thanks to a technological revolution in solar power, wind power, and, less visibly, batteries.
So let me take an optimism break and talk about why batteries may save the world.
The decline in battery prices has been incredible. It’s like nothing anyone has ever seen before. Big, strong men with tears in their eyes come up to me and say, “Sir, have you seen the progress in batteries?”:
Why does this matter?
First, cheap battery storage of electricity greatly mitigates the problem of intermittency — the sun doesn’t always shine, the wind doesn’t always blow. This was a major concern early in the renewable revolution. Some energy economists scolded me for my naïve optimism when I first wrote about solar technology way back in 2011. But solar + batteries provides round-the-clock power.
Here’s a graph of California’s electricity supply generated by renewables and batteries over the course of 24 hours on April 1 that illustrates my point:
During the middle of the day, California generates lots of electricity from solar. Much of it is poured into batteries, which provide electricity when the sun sets. Californians don’t even notice the switch.
Second, battery performance has soared as prices have plunged. Crucially, there has been a huge increase in batteries’ volumetric energy density: the amount of electricity that can be stored in a given space. Until a few years ago the energy density of gasoline gave internal combustion a huge advantage over electric vehicles. But no longer. Outside the U.S. electrification, the transition away from petroleum and towards electricity — particularly green-sourced electricity — is well underway:
Third, we should expect continuing rapid improvement in renewable energy. That’s because the progress in batteries has come from cumulative learning rather than scientific breakthroughs. Lithium-ion batteries are, in fact, a decades-old technology. Yet costs have fallen drastically and energy density risen thanks to an ongoing process of learning, which shows no sign of coming to an end.
Furthermore, we’ve seen rapid progress in all components of the green energy transformation, even though their underlying technologies have little in common. Solar panels, wind turbines, and batteries are very different, yet all have seen revolutionary improvements. This strongly suggests that the whole renewable energy complex is experiencing a virtuous circle: ever-growing use leads to falling costs and falling costs lead to ever-growing use.
If we ask where this virtuous circle is taking place, the answer is, largely in China with an assist from Europe. And the corollary is “not in America.” The United States has allowed itself to be far surpassed by China and is now only a peripheral player in the renewable revolution. Fortunately for the rest of the world, this means that the Trump administration’s hostility to renewable energy, its attempts to sabotage progress, won’t stop that revolution or even noticeably slow its momentum. True, Trump’s anti-green, pro-pollution tilt will serve to leave America further behind, but progress in fighting climate change and reducing the risks of global dependence on oil will continue.
So although we are now in the midst of a severe energy crisis that could easily go on for many months, this too shall pass. A better, cheaper, cleaner energy future is on the way, and not even Trump can stop it.
MUSICAL CODA
Trump is idolized because he re-opened the civil war, that the butthurt southern states lost but have been still fighting in cosplay for the last century and a half. It never actually ended in their minds. Time to finally nail that coffin shut.
God, finally something to smile about!
No posts

source

Ready to get up to 3 free quotes?
Get up to 3 free quotes for solar, batteries, EV chargers or hot water heat pumps
GET MY QUOTES

Despite showing up at All Energy conference with a perfectly swank and expensive trade stand, it appears SMA Australia has quietly withdrawn from the Australian market.
Surely SMA haven’t gone broke? No, they’ll maintain a presence in the large scale commercial sector for solar farms.
It’s been a couple weeks since I saw a grainy photo of a PC screen which put a ripple of disbelief through solar circles, however the PDF copy obtained since is clear enough. What was once the undisputed Australian market leader in solar inverters, SMA have packed up their sales operation completely.
This is all the confirmation we have at the moment.
A foundational part of mass market solar and ongoing part of the industry furniture, SMA will be missed by the kind of people who prioritised quality and longevity.
Probably the oldest SMA I encountered – a BP Solar branded GCI200 coupled to Australian made, 75 watt frameless BP Solar panels.
Thankfully SMA are honourable enough to honour their warranties – they’re apparently maintaining a local office so everyone with a SunnyBoy/Sunny Island/Sunny Storage can still enjoy a bright outlook. Although I don’t know how “smart connected” services will work going forward.
It’s a stark contrast to Hanwha effectively abandoning Australia when they pulled the pin on Qcells.
As a sole trader I never advertised. Either I got to yarning with people or my phone simply rang because someone had recommended me. For years I simply installed SMA Sunny Boy TL5000 inverters with 20 plus panels on the roof, and they were rock solid.
Reliability personified.
The only time I weakened, a customer talked me into a cheap piece of junk, which taught me a great lesson. You should never compromise your standards because replacing a Growatt 5 times over doesn’t pay.
However none of my customers have ever rang back to complain about the stout red box on the wall, but sadly the sands of time have caught up with some of the Simax panels I’d installed with SMA inverters. The guys at Suntrix said Simax were excellent quality, but in retrospect I should have been selling REC panels.
When your panels turn out to be rubbish and the water leaks into the edges, you end up with earth faults which knobble output until they possibly dry out. The problem will only get worse and your SMA inverter will protest with a red light and an isolation error on the screen.
“Insulation resist” and the dreaded red light have brough production to a stop after 91.926MWh and 11yrs 4 days – roughly 22.85kWh/day. About the only flaw to report was an occasional screen failure simply due to age.
As SolarQuotes founder Finn Peacock commented to me about SMA recently: “they did it to themselves”. It’s a shame really, but when you’re leading the market there’s always a possibility of falling off the wheel.
As I recall, there were a few factors which may have brought SMA undone. A 2008 world economic crisis was dodged by Australia, but the German company didn’t maintain production enough to satisfy the burgeoning market here. My own house ended up with an Australian made Latronics PVE2500 because we couldn’t buy anything else.
The incredibly heavy and robust SunnyBoy 1100, 1700 & 2500, or SMC series were the industry standard for many years, but when SMA moved to transformerless topology the TL 3000, 4000 & 5000 took over everywhere. Then came the HF units for a short while.
The SMA HF3000 solar inverter.
However the real defining moment was when the German-manufactured Sunny Boy TL was superseded around 2016 by the AV 40. All of a sudden we had “premium” products that were dead on arrival. Installers were already upset that the screen had gone missing, but a ludicrous quality control failure that delivered brand new but broken inverters just torched SMA’s reputation.
I’ve never seen an AV40 catch fire but they certainly incinerated SMA’s reputation.
Everyone said screw you and your move to Chinese manufacture. Especially when there was a separate cheap brand brought out with SMA support. ZeverSolar had a short life and I’m thankful I only ever dirtied my hands on one of them.
SMA Sunny Island 48V battery inverters turned up everywhere, including this Redfow off grid system with a tonne of lead batteries in the back end.
When Fronius came out with the snapinverter range, the rest was history. While SMA had replaced the trusty and infomative LCD screen with 3 LEDs and a newfangled monitoring app, they found people just don’t like change.
Fronius had an equally good Austrian reputation and they had a better screen. With the right code installers had probably a hundred menus accessible via 4 buttons.  Solarweb online monitoring available via WiFi and no pesky bluetooth interface, it was a real winner.
It seems SMA have just lost interest in Australia. Even with the release of the new hybrid battery systems in October 2023, the EV charger wasn’t part of the Australian lineup, Though we have at least one 5 star review of it being installed.
As recently as March 2025 they were talking up a recovery after some pretty ordinary results, but it seems Australia just isn’t part of the plan.

Please leave us some comments, or better still, write a review if you have a good yarn to tell about SMA. You never know, they might come back one day.
Sign up for our weekly newsletter!
Your emailSUBSCRIBE
Anthony joined the SolarQuotes team in 2022. He’s a licensed electrician, builder, roofer and solar installer who for 14 years did jobs all over SA – residential, commercial, on-grid and off-grid. A true enthusiast with a skillset the typical solar installer might not have, his blogs are typically deep dives that draw on his decades of experience in the industry to educate and entertain. Read Anthony’s full bio.
They sound like the “Nokia” of the solar industry. There are plenty of old time installers who wax lyrical about the sunny boys, never heard one talk up anything newer from the company.
Surprised they stayed 10 years after effectively killing their product.
With the passing of the king of inverters, it begs the questions,
1. who is the successor or has the kingdom been divided between the Lords of Europe and Asia.
2. how long will they reign, hopefully much longer than the warranty period?
3. what can be learned from the king’s passing so that history does not repeat itself. SMA removed its inverter screen. What brilliant but ill-conceived idea will poison the new reign. Will it be something to do with AI, bluetooth, ethernet or VPP connection? Or is the writing on the wall for lithium?
4. what is happening to the courtiers (employees) of Australia’s SMA empire? Is SMA looking after its staff who liased with us solar peasants in the sale and maintenance of the ubiquitos red and blue boxes? Did they arrive at work to find their front door entry code no longer worked and a sign saying “the personals from your desk will be posted to your home address.”
Feel free to weigh-in. Probably, only the last question is the most pressing.
The three basic rules that apply to any business entity are the cost to get in, the cost to stay in, and the cost to get out.
Today, it is less about the hardware and more about the functionality and the interface, which can be accessed [reporting] and/or configured by the user.
I can access my solar and storage systems on my iPhone from any location with internet access. If you are not in the cloud you have nothing to offer.
Also data reliability. If you get garbage output every daymonthyear you have a blackout then the information you see will be useless, unless you actually believe your house managed GWh output one day of the year, or achieved negative output on another!
My data recorded by the retailer, inverter solar and battery are all within acceptable tolerances of each other.
Your new-fangled cloud-whatever system will suffer from its own form of technology rot way way faster than a sheet of unpainted mild steel if left on the tidal zone on a beach in a tropical area.
Still, enjoy it while you have it.
I guess SMA cost-cut themselves out of contention a long time ago.
The days of anything stamped “.. In Germany” with its implied good product design and whatever standing are long gone especially when everything is a hodge podge of bits and pieces from half a dozen or more places with dubious QC.
And then consider your shiny new Internet everywhere connected kit is only ever as good as the weakest link. And so much of these weak links are often software – usually in the form of crappy software locked inside Bluetooth or wifi modules or other embedded components that simply can’t be updated in the field. That’s the start of the rot right there.
In top of that add the cloud based systems needed to make all that work cost a ton of money to build and run.
I have an SMA Sunnyboy SB1100 Inverter, quietly doing the business since March 2008 without problems, producing some 20,066kWh worth of green electrons and STILL going strong, a testament to ‘Made in Germany’ quality.
Their greed was a contributor as well. No more Primo’s. In stead, you had to buy an expensive GEN 24 with fan failures. Deye was such a cheaper option and with SMA staying power.
I have 3.3Kw of REC Solar Panels paired with SMA TL4000 which ahs been running flawlessly since April 2011. Over 65,000Kwh produced in that time in Geelong VIC
Staff probably saw the recent industry growth and left leaving SMA with fewer staff and they found it hard to attract people. A dying star.
This is like the former car industry. And it will be what happens to the US car giants. Chinese supply with lower costs and improved tech will kill the noble. Cost competition destroys originators and those that sell at high prices by selling fear. Just as japanese cars were surpassed by korean cars and now china dominates. No us car giant can recover. They wind back and collapse. Not just in production scale but with tech development and improvements.
All things solar are headed that way. Fight it or flight it ! So many sales pitches for anything solar start and end with fear of cheap chinese quality. A few years ago you couldn’t give away some cars made in china. I see the solar industry heading this way..
I had a great run with an SMA TL5000 inverter. The company went quietly belly up in 2016 or 2017, which wasn’t surprising considering the abysmal initial install job. Something caused all 14 panels to get hot spots in the middle going into bypass and robbing the system of much needed voltage output. But it soldiered on until a few years ago when something fell out of the sky and totaled one panel, which tried to catch fire and burn my shed. It is current off line until I can find one or more panels and get it going again. I was glad that I insisted on the SMA Inverter which still should be capable of going back into service. One right choice out of 3 with the Hyundai panels with no warranty as the dodgy company direct imported them, and the crap install finally fixed a month after the initial switch on. approved by an electrical inspector who never visited the property and a supervising electrician who was never on site for the installation..
Sounds like you had a crappy time for the original install – a good example of the reasons we have the current regime of daily limits and photographic proof for the current process for installing systems!
I got 20 Les of solar installed in 2012(10 on the house and shed) using SMA inverters here in Melbourne.Still going strong
The end was nigh when SMA handed production to the Chinese. An inevitable result is my opinion,as they put profit over good performance.
Tried a couple of times about wanting to buy and connect a 20kwh sodium ion battery. Found the battery they don’t call back of even reply via email. Try and find a company to install a sodium ion battery, yeah I get they are not up to that technology. Sorry, but it is already here and CATL the largest battery producer in China has already developed Sodium Ion batteries for EV’s. Many installers say they are also booked till September. What I also want is to add an additional 4kw of solar panels to my existing array. So here we are at a dead end with an additional 4kw of solar panels and a 20kwh sodium ion battery and not an installer to be had. So much for the green dream.
There are no sodium batteries approved for residential use as yet in Australia – to the best of my knowledge.
So you might have to wait a while if that’s what you want. Unless you are an “early adopter” willing to deal with any teething problems, it might also be wise to wait a while even after they are approved.
Please keep the SolarQuotes blog constructive and useful with these 5 rules:
1. Real names are preferred – you should be happy to put your name to your comments.
2. Put down your weapons.
3. Assume positive intention.
4. If you are in the solar industry – try to get to the truth, not the sale.
5. Please stay on topic.



Notify me of follow-up comments by email.
Notify me of new posts by email.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Δ
 RSS – Posts
Read The Good Solar Guide Free Online!
Ready to get up to 3 free quotes?
Get up to 3 free quotes for solar, batteries, EV chargers or hot water heat pumps
GET MY QUOTES

© 2009 to 2026 SolarQuotes Home Electrification Pty Ltd
Get free quotes for solar, batteries,
EV chargers or hot water heat pumps

Download the first chapter of The Good Solar Guide, authored by SolarQuotes founder Finn Peacock, FREE!

You’ll also start receiving the SolarQuotes weekly newsletter, keeping you up to date on all the latest developments on Australia’s solar scene.
We respect your privacy and you can opt out from the newsletter at any time.

source

You’ve finally got the full trifecta – a rooftop solar panel system feeding a home backup battery that charges your EV. Heck, you’ve even got a bidirectional EV that can send power back to the house when you need it. Now the real question: are you actually saving any money?
Last week, microgrid expert and GM Energy employee Jim Reilly took to LinkedIn to share his personal experience with a whole home electrification setup that includes both a GM Energy home backup battery and a GMC Sierra EV powered by electrons generated by his rooftop solar array – and he’s coming out way ahead.
“Lots of discussions on how tensions in Iran threaten a gas price surge,” writes Reilly, in his original “Energy Dominance” post. “The cost of a ‘full tank’ of gas is unpredictable. Here’s my math on a ‘full tank’ so I can check in on this in a few months, to see if these numbers spread even further, driving more people to an electric vehicle.”
⛽ Gas (24 gal at $4/gal): $96 per fill-up.
🔌 Public EV (200 kWh at $0.48/kWh): $96.00 per fill-up.
🏠 Home EV (PSE at $0.14/kWh): $28.00 per fill-up.
☀️ Solar EV (if sized larger correctly for your home): $0.00 (Locked-in independence).
JIM REILLY
Juicing up the massive 205 kWh Max Range battery in the GMC Sierra EV for less than $30 is a game changer in a world with $5 gas and $7 diesel – especially in light of the the truck platform’s record-setting range, and the fact that gas and diesel prices don’t seem to be coming down anytime soon.
As Reilly points out, the real benefit of pairing an EV with a home solar system is the ability to make your own fuel, and help set the price of refueling by selecting when you want to charge, when you can, and storing the most affordable energy in your batteries for use later. And, for his part, he absolutely gets it.
Head over to Jim Reilly’s LinkedIn account and follow along for more solar and fuel cost comparisons, as well as more of his work compiling and analyzing his home solar data and energy savings, then let us know if you’ve got your own, similar data to share with us in the comments.
If you’re curious how this math plays out in the weeks ahead, Reilly says he’ll keep documenting his home setup’s performance and sharing more details and cost comparisons on his LinkedIn account. Accounts like his are great because they’re not just theory – they’re real-world examples of how a whole home setup like Jim’s can work.
While Reilly’s setup seems to be working incredibly well for him, your home, battery, EV, and driving habits might be dramatically different. If you’re considering following in his footsteps, talking to a qualified professional installer can help you understand what’s being offered and how a given deal is being structured. Take the information they give you to an accountant or trusted financial expert to understand what’s real, what’s marketing, and what actually makes sense for you — and if there’s money on the table in the form of local utility incentives or tax credits, make sure you don’t leave it there.
Unless, you know, you don’t actually care about money.
YOU MIGHT ALSO LIKE:
If you’re considering going solar, it’s always a good idea to get quotes from a few installers. To make sure you find a trusted, reliable solar installer near you that offers competitive pricing, check out EnergySage, a free service that makes it easy for you to go solar. It has hundreds of pre-vetted solar installers competing for your business, ensuring you get high-quality solutions and save 20-30% compared to going it alone. Plus, it’s free to use, and you won’t get sales calls until you select an installer and share your phone number with them. 
Your personalized solar quotes are easy to compare online and you’ll get access to unbiased Energy Advisors to help you every step of the way. Get started here.
FTC: We use income earning auto affiliate links. More.
Subscribe to Electrek on YouTube for exclusive videos and subscribe to the podcast.
I’ve been in and around the auto industry for over thirty years, and have written for a number of well-known outlets like CleanTechnica, Popular Mechanics, the Truth About Cars, and more. You can catch me at Electrek Daily’s Quick Charge, The Heavy Equipment Podcast, or chasing my kids around Oak Park, IL
Find a reliable home solar and battery installer and save 20-30% compared to going it alone.
Qmerit makes electrification easy — connecting you with trusted pros who get it done right.

source

Japanese researchers developed a molybdenum-based spin-flip emitter that efficiently harvests triplet excitons from singlet-fission tetracene dimers, producing strong near-infrared emission. This approach could boost solar cell efficiency and enable new quantum technologies by converting otherwise “dark” excitons into usable light.
Researchers successfully capture singlet-fission–amplified excitons with a molybdenum-based emitter, achieving 130% quantum yield and opening a path beyond solar cell efficiency limits.
Image: Kyushu University
A research team at Kyushu University in Japan has reported a breakthrough that could steer photovoltaic technology past long‑standing efficiency barriers by harnessing a quantum process known as singlet fission (SF).
Singlet exciton fission is an effect seen in certain materials whereby a single photon can generate two electron-hole pairs as it is absorbed into a solar cell rather than the usual one. The effect has been observed by scientists as far back as the 1970s and though it has become an important area of research for some of the world’s leading institutes over the past decade; translating the effect into a viable solar cell has proved complex.
Singlet fission solar cells can produce two electrons from one photon, making the cell more efficient. This happens through a quantum mechanical process where one singlet exciton (an electron-hole pair) is split into two triplet excitons. By pairing SF with a specially designed spin‑flip molybdenum‑based complex, the scientists demonstrated energy conversion and harvesting in solution with an effective quantum yield of around 130%.
“The applications of this work in solar cells will require integrating singlet-fission (SF) materials with spin-flip emitters in solid-state systems,” Nobuo Kimizuka, lead author of the study, told pv magazine. “As fundamental research, our first step is to develop high-efficiency SF and spin-flip emitters with well-controlled energy levels and luminescence quantum yields in solid-state environments, and then evaluate the performance of these integrated systems.”
“We are actively working on building a higher-performance solid-state system,” he added. “Achieving robust performance in solid-state solar cells remains a challenge, but we expect the efficiency to surpass that of conventional SF technology alone. This approach, which multiplies photons and converts otherwise ‘dark’ triplet excitons into light, could open the door to new quantum technologies such as quantum sensors and exciton circuits, while also contributing to the design of next-generation quantum materials.”
The team developed a molybdenum-based spin-flip emitter that selectively captures the energy of triplet excitons before they dissipate. Its molecular design allows electron spin to flip during near-infrared (NIR) light absorption or emission, enabling more efficient harvesting of the multiple excitons generated by singlet fission.
Further analysis showed that sensitization efficiency depends heavily on the structure of the linker connecting tetracene units. The linker dictates not only the spatial arrangement and electronic coupling of the chromophores but also the exchange interaction within the correlated triplet pair. Variations in linker length, rigidity, and conjugation can significantly affect the rate and yield of triplet energy transfer to the spin-flip emitter, influencing both efficiency and the dynamics of the singlet fission process.
“The methodology we developed for assessing doublet yields provides a practical way to estimate triplet yields of SF dimers, even in systems with complex energy-transfer pathways involving both correlated and free triplets,” Kimizuka explained. “Reducing losses from correlated triplet-pair recombination requires either rapid separation into long-lived multiexcitons or faster triplet transfer to an acceptor molecule, achievable through careful energy-level design in oligomers or solid-state structures.”
“With a versatile selection of central metals, including chromium, molybdenum, and vanadium, and tunable ligands informed by Tanabe–Sugano diagrams and ligand-field theory, spin-flip emitters show strong potential as NIR-emitting materials for efficient triplet extraction, especially with recent advances in air-stable designs,” he added.
The interface design will be critical for converting triplet excitons generated by tetracene singlet fission into charge carriers on the silicon solar cell surface. “In SF-sensitized silicon cells, one major source of energy loss is transfer from the SF molecule to silicon via its excited singlet state,” Kimizuka noted. “Our proof-of-concept method blocks these loss pathways, enabling selective extraction of the excited triplet states originating from singlet fission.”
The research findings are available in the study “Exploring Spin-State Selective Harvesting Pathways from Singlet Fission Dimers to a Near-Infrared-Emissive Spin-Flip Emitter,” published in the Journal of the Chemical American Society.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Emiliano Bellini
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

DARIEN — A field full of solar panels in rural Walworth County could soon sprout a natural gas-fueled power plant. The planned Foundry Ridge Energy Center is now in the environmental review stage.
The proposed power plant near Darien from energy developer Invenergy aims to provide 324 megawatts of power during peak energy usage times. Some neighbors who live next to the proposed site are not happy about the potential development.
Watch: Darien neighbors worried about planned power plant:
“It’s difficult when you choose how you want to raise your family and this gets slammed at us,” said Cheryl Simer, who lives within walking distance of the potential Foundry Ridge site. “Don’t put this in our yards. Don’t put this within a half a mile of anyone’s home.”
“We don’t want to see this monstrosity in our community,” said Hannah Schlick, who lives across Turtle Creek from the solar fields and Foundry Ridge site. “We don’t want to deal with the potential impact of our air quality, our wells, and our water quality.”
TMJ4 asked Invenergy what the impacts of the project would be if it is approved by the Public Service Commission of Wisconsin.
In a statement, they said:
The statement continued:
The Public Service Commission is taking comments from the public on the Foundry Ridge proposal until April 24. You can share your feedback by clicking here.

It’s about time to watch on your time. Stream local news and weather 24/7 by searching for “TMJ4” on your device.
Available for download on Roku, Apple TV, Amazon Fire TV, and more.

Report a typo or error // Submit a news tip
Report a typo

source

LEXINGTON, Ky. (WKYT) – The Lexington City Council is taking another look Thursday night at a ground lease for a solar farm on the Haley Pike landfill after tabling the item at its last meeting.
The Lexington Model Airplane Club, which sits on the Haley Pike landfill, could be kicked out by a 357-acre solar farm. Lexington’s City Council postponed a lease agreement to Edelen Renewables earlier this month, but it is back on the table.
A former member of the airplane club and STEM leader in Lexington is proposing sharing the land.
“It’s worth giving up a little bit of solar energy to keep this around,” said Keith Hollifield, chief toy maker for Newton’s Attic.
Hollifield created a plan for a smaller solar farm, which would allow the airfield to stay, while also partnering with Newton’s Attic to build a STEM park.
Newton’s Attic is a STEM education nonprofit that offers hands-on project-based learning. The organization has been providing STEM education resources in Lexington for 27 years, but is limited by ground and air space at its current location near Blue Grass Airport.
“This is the perfect place to do it because we can literally hit the ground running; it is ready to go today,” Hollifield said.
Hollifield said the 680-acre former landfill has room for more than just a solar farm.
“We need to be careful what we take away in advancing these other initiatives, and I think what we’re giving up here is simply too much. A slightly smaller solar farm, plus this is way better, infinitely better,” Hollifield said.
He hopes the council will at least find a solar company willing to compromise.
“No one has said that if you don’t figure it out, you can’t do the project. No one has really put their feet to the fire, that’s what I’m asking the council to do: essentially say that,” Hollifield said.
Leaders at Edelen Renewables state that they have been collaborating with the Lexington-Fayette Urban County Government on this proposal since July 2025. COO Amy Samples said when the city opened the request for proposal, they marked all 357 acres as available.
“With the proposal having been marked as ‘land available for solar development,’ our proposal did respond to that,” Samples said.
Samples said even with two weeks to reconsider a shared-use project after the council tabled the decision, she said it’s not feasible. She cited overall costs, insurance concerns, and optimal panel layout.
“Even if our project did scale based on our outstanding results from the Kentucky Utilities Project, this parcel remains pretty integral to the success of this project,” Samples said.
The LFUCG Council meeting starts at 6 p.m. Thursday at the Government Center.
Copyright 2026 WKYT. All rights reserved.

source

A German research team has developed CuInSe₂ micro-concentrator solar cells using laser-assisted metal-organic chemical vapor deposition to grow indium islands directly on molybdenum-coated glass, forming absorber arrays without masks or patterning. The not-yet-optimized micro-modules achieved up to 0.65% efficiency under one sun, with gains of up to 250% under concentrated illumination.
Schematic of the CuInSe2 micro-solar cell
Image: Universität Duisburg-Essen (UDE), Solar Energy Materials and Solar Cells, CC BY 4.0
A research team in Germany has developed a copper, indium, selenium (CuInSe₂) micro-concentrator solar device composed of vertically grown absorber islands on a molybdenum (Mo) films.
The scientists used laser-assisted metal-organic chemical vapor deposition (LA-MOCVD) to grow indium (In) islands in a bottom-up approach, instead of depositing a continuous thin film and subsequently patterning it. “The primary novelty of our work is the use of a LA-MOCVD method for the bottom-up growth of indium precursor islands,” corresponding author Jan Berger told pv magazine. “This approach proved to be a fast and reliable technique for simultaneous local growth, importantly offering the possibility to add gallium and copper locally using the same method.”
“The most unexpected finding was that the indium precursor islands formed distinct cluster structures that remained pinned in place, refusing to coalesce into a single large island – even after annealing above the melting temperature of indium,” he added. “Furthermore, it was surprising to see that the structural features of these precursor islands remained clearly visible even after the selenization process.”
Device fabrication begins with glass substrates coated with Mo, which are then processed by LA-MOCVD. In this step, a laser array locally heats the substrate. It decomposes the precursor gas only at defined spots, forming a 7 × 7 array of indium islands without the need for masks or patterning. A thin copper layer is subsequently deposited, and the stack is selenized to form CuInSe₂ absorber islands.
Image: Universität Duisburg-Essen (UDE), Solar Energy Materials and Solar Cells, CC BY 4.0
Afterward, the samples are etched to remove unwanted material, coated with photoresist for electrical isolation, and patterned with a laser to form openings. The solar cell is then completed by depositing a cadmium sulfide (CdS) buffer layer, followed by intrinsic zinc oxide (i-ZnO) and aluminum-doped zinc oxide (AZO) window layers. Finally, each array of 49 micro-cells is contacted and measured as a single module, with a device structure of glass/Mo/CIS absorber/ cadmium sulfide (CdS)/i-ZnO/AZO.
Overall, the team produced nine micro-modules and tested four of them. Initial measurements were conducted under one sun, followed by increasing intensities up to 17 suns to simulate concentrator conditions. These not-yet-optimized arrays achieved a conversion efficiency of up to 0.65% under one sun, with efficiency rising under higher illumination—gains of around 60% at lower concentrations and up to 250% at 17 suns.
“Functional devices were successfully produced, but notable key challenges were identified, particularly related to the intensity distribution of diffractive optical element (DOE), the initial morphology of indium islands, and process repeatability. Addressing these challenges in terms of material quality and process control is essential,” the team explained. “Once resolved, the LA-MOCVD method holds significant promise as a rapid and resource-efficient production technique for next-generation micro-concentrator photovoltaics.”
The new cell concept was presented in “CuInSe2-based micro-concentrator solar cells fabricated from In islands grown by laser-assisted MO-CVD,” published in Solar Energy Materials and Solar Cells. Scientists from Germany’s University of Duisburg-Essen, the Leibniz Institute for Crystal Growth, the Federal Institute for Materials Research and Testing, Brandenburg University of Technology Cottbus-Senftenberg, and the engineering company Bestec have participated in the study.
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
More articles from Lior Kahana
Please be mindful of our community standards.
Your email address will not be published. Required fields are marked *
Comment *
Name *
Email *
Website
Save my name, email, and website in this browser for the next time I comment.


Δ
By submitting this form you agree to pv magazine using your data for the purposes of publishing your comment.
Your personal data will only be disclosed or otherwise transmitted to third parties for the purposes of spam filtering or if this is necessary for technical maintenance of the website. Any other transfer to third parties will not take place unless this is justified on the basis of applicable data protection regulations or if pv magazine is legally obliged to do so.
You may revoke this consent at any time with effect for the future, in which case your personal data will be deleted immediately. Otherwise, your data will be deleted if pv magazine has processed your request or the purpose of data storage is fulfilled.
Further information on data privacy can be found in our Data Protection Policy.
Legal Notice Terms and Conditions Data Privacy © pv magazine 2026
Δ
This website uses cookies to anonymously count visitor numbers. View our privacy policy. ×
The cookie settings on this website are set to “allow cookies” to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click “Accept” below then you are consenting to this.
Close

source

World’s Only Pure Car & Truck Carrier With The Largest Photovoltaic System  marineinsight.com
source

TOLEDO, Ohio (WTVG) – People living in Toledo’s Old South End are worried about what will happen if their greenery gets replaced by a proposed solar field.
Neighbors say the site a developer is eyeing is filled with trees, plants, and animals. They’re concerned about the effects a solar field might have on the wildlife.
Rachel Shoup says she walks with her children along Lotus Avenue every day. It’s on their route to school, but they also often stop and admire the wildlife in the field that sits in between Lotus Avenue and the Maumee River.
“They see wild turkeys down there, they see deer down there, we’ve eagles, foxes, rabbits, you name it, they’re down there,” Shoup said.
When Shoup and her family heard that the Toledo-Lucas County Plan Commission was hearing a case on whether to approve building a solar field in that space, they were frustrated.
“It won’t be as pretty and there won’t be as much nature,” said Zoey, Shoup’s 13-year-old daughter.
Other neighbors living on Lotus Avenue had similar thoughts about the wildlife.
“My concern is a lot for the migrating birds,” said Penny Noyes, who lives on Lotus Avenue.
Noyes questioned whether solar panels may confuse birds flying overhead, who may mistake them for the river.
The Historic South Initiative, a nonprofit that fixes up homes in the Old South End, is the one planning the solar field. In its proposal to the plan commission, leaders wrote that they want to use 4.6 acres for a standalone solar field.
The Historic South Initiative would sell the energy at a reduced price. That money would fund the group’s improvement projects.
Sue Terrill grew up right next to the greenery near Lotus Avenue.
“We’re not against the solar panels, we’re against the location against the river and for the neighbors,” Terrill said.
Her concerns about the proposed solar field led her to create fliers about the situation. She’s also spreading the word about a meeting with residents and the Historic South Initiative. It’s set for Monday, April 6 at 6 p.m. at the South Branch Library.
The Toledo-Lucas County Plan Commission pushed back its vote on the case until after the Historic South Initiative meets with residents.
The Plan Commission is expected to vote on the case during its meeting on April 9.
Latest Local News | First Alert Weather | Crime | National | 13abc Originals
Copyright 2026 WTVG. All rights reserved.

source

source

eNewspaper
Sign up for email newsletters

Sign up for email newsletters
eNewspaper
Trending:
Copyright 2026 Chicago Tribune. All rights reserved. The use of any content on this website for the purpose of training artificial intelligence systems, algorithms, machine learning models, text and data mining, or similar use is strictly prohibited without explicit written consent.

source

Nick Gier in “Opinion: Solar energy surges in developing nations” (Daily News, March 17/19) states that the scrapping of the 2009 EPA endangerment finding does “not address the overwhelming evidence against the dangers of CO2 emissions.”
I would encourage him to do some research on CO2. We all breathe out several pounds of CO2 every day. CO2 in ambient air is not a health problem. It’s not visible and is vital for plant growth. Without significant CO2 in our air, we would not have food to live on. Plants for the most part would thrive even better if the percent of CO2 in the air were several times higher.
The air pollution Gier talked about in his column for the most part was not CO2 but particulate caused by a number of primitive activities such as cooking and heat over an open fire, using animal dung as fuel, using engines that did a poor job of combusting the fuel, poor farm practices and more. This kind of air pollution is cleaned up by developing nations as they are able to afford cleaner, more efficient practices of burning fuel of all kinds and having funds to spend on upgrading various primitive operations from farming practices to mining and home heating. Green energy programs, when done efficiently, are one way to reduce air pollution but more efficient combustion of fossil fuels is also a way, which is economically and environmentally more efficient, especially in those parts of the Earth that do not consistently get direct sunlight and adequate daily wind. Hydropower is a dependable green energy that is not included in what is said here.
The biggest problems with wind and solar energy are: 1) the environmental problems associated with mining for the basic materials to make the panels and turbines, 2) the intermittent aspect of these energy sources, 3) the huge amounts of land required, 4) they are relatively short-lived and 5) associated environmental drawbacks such as waste disposal and the killing of relatively large numbers of birds in their operation.
I am not a climate change denier but I am a realist. I would love to see green energy become economically viable but until we can harvest solar energy in space and transmit it cost-efficiently 24/7 to Earth, green energy, whether solar or wind, is going to be a niche energy source that is economically and environmentally viable only in limited cases.
The United States and some other developed countries have spent trillions of dollars attempting to replace conventional energy sources with green energy. Worldwide there has been limited success because of the drawbacks mentioned above. Since wind and solar energy harvested on Earth is intermittent, power companies must have backup energy to supply the needed energy when there is no sun and the wind is not blowing. This typically means having fossil fuel or nuclear power plants as the backup energy source since we have not yet developed batteries that can economically handle the needed storage. Thus power companies are forced to pay for both energy sources, which is far more expensive than either one alone and raises rates for all. In addition, the energy needed for mining and manufacturing solar panels and wind turbines, which is significant, at this time must be drawn from fossil fuels.
As a result of these drawbacks, less than 10% of U.S. energy is from solar and wind at this time, in spite of the huge investments toward conversion. Europe, especially Germany, has also made huge expenditures into green energy with negative success, resulting in large increases in the cost of public energy in general and significant weakening of their economies. California in the U.S. is a good case study of the failure of green energy to live up to expectations. I am all for technology that will help clean up and maintain a healthy environment but it will not happen unless it is economically viable in either developing or developed countries.
To my knowledge, none of the large AI facilities being built or planned, which require huge amounts of energy, will be powered even in part by green energy. This is because the energy source for these facilities must be dependable 24/7 and green energy is definitely not at this time.
I encourage Nick Gier and others to keep pushing research for green energy solutions but to recognize the truth that green energy is only a niche solution at this time and it will not allow developing countries to achieve developed status.
Kirkland is a retired environmental engineer and businessman, spending time at the University of Idaho, Washington State University and the Moscow Recycling Center. He also participated on the Moscow City Council and was an elder at various churches for more than 40 years.

source

Top Stories Of The Day: Cosmic PV Files ₹640 Cr IPO; Neuron Launches BESS Facility and More…  SolarQuarter
source

© 2025 THE COOL DOWN COMPANY. All Rights Reserved. Do not sell or share my personal information. Reach us at hello@thecooldown.com.
“We cut everything we didn’t need … to make the molecule as compact as possible.”
Photo Credit: iStock
Researchers at the University of California, Santa Barbara, have made a solar battery that can store and discharge heat energy under sunlight, setting the stage for a new kind of power generation and storage system without panels. 
As doctoral student and lead study author Han Nguyen explained in a news release, the process is similar to photochromic sunglasses, which adapt to light conditions. 
“That kind of reversible change is what we’re interested in,” Nguyen added. “Only instead of changing color, we want to use the same idea to store energy, release it when we need it, and then reuse the material over and over.”
The secret is chemical bonds that can discharge heat on demand as part of a “liquid” battery, which can store energy for years. It’s an impressive device with double the energy density — the amount of electricity stored per pound — of common lithium-ion packs. 
If successfully developed, the innovation could provide an energy storage solution without arrays, heavy batteries, or the electric grid. The reusable and recyclable invention was referenced as a “concept” in the report. 
To unlock the ability, researchers are using a modified organic molecule called pyrimidone, which is structurally similar to a component in DNA. When exposed to ultraviolet light, it changes structure as part of a reversible process. The team examined how the molecule works and why it can remain stable for so long.
Pyrimidone is springlike when hit with sunlight, and it twists into a “high-energy shape.” It remains locked until a catalyst triggers it to snap back, releasing heat energy. 
“We prioritized a lightweight, compact molecule design,” Nguyen said in the release. “We cut everything we didn’t need … to make the molecule as compact as possible.”
The prototype passed a crucial benchmark by boiling water. This opens the door for various heating applications at off-grid sites or in residential buildings. It’s an exciting entry in the development log for molecular solar thermal research, according to the scientists. 
“Boiling water is an energy-intensive process,” Nguyen said. “The fact that we can boil water under ambient conditions is a big achievement.”
💡These best-sellers from Quince deliver affordable, sustainable luxury for all
Starting at $50
Starting at $99
Starting at $60
Starting at $80
Which of these savings plans for rooftop solar panels would be most appealing for you?
Save $1,000 this year 💸
Save less this year but $20k in 10 years 💰
Save less in 10 years but $80k in 20 years 🤑
Couldn’t pay me to go solar 😒
Click your choice to see results and earn rewards to spend on home upgrades.

Solar energy via panels is already a great way to leverage the sun’s rays to counter surging electricity rates and a grid that’s reliant on expensive oil, gas, and coal — which all contribute to poor air quality. If you’re interested in going solar, here are some tools to get you started.
• EnergySage can help you save up to $10,000 on installations by curating competitive bids from local installers
• Not ready to spend up front? Palmetto’s $0-down LightReach solar leasing program can lower your utility rate by up to 20%
• TCD’s Solar Explorer makes it easy to access exclusive offers from preferred partners
Pairing solar panels with energy-efficient tech is a great way to augment the savings. Upgrading aging HVAC systems is a great place to start, and TCD’s partner Mitsubishi can help you find the right heat pump for your home at the right price. 
The Palmetto Home app is another great resource that unlocks up to $5,000 in rewards to spend on home upgrades, earned by taking some simple, smart actions in everyday life.
Get TCD’s free newsletters for easy tips to save more, waste less, and make smarter choices — and earn up to $5,000 toward clean upgrades in TCD’s exclusive Rewards Club.
© 2025 THE COOL DOWN COMPANY. All Rights Reserved. Do not sell or share my personal information. Reach us at hello@thecooldown.com.

source

RIESEL, Texas (KWTX) – A nonprofit is partnering with solar farms across the nation to set up bee hives on small plots of land, providing therapy for veterans and first responders while supporting pollinator populations.
At the Roseland Solar Farm in Riesel veterans and first responders are paired with experienced beekeepers to care for the bee hives on the site and harvest their honey.
“Our mission at Hives for Heroes is to save bees and save lives,” said Steve Jimenez, president and founder of Hives for Heroes.
Jimenez, a veteran himself, explains that he wanted to find a way to help others who may be struggling.
“We have an extremely high suicide rate, you know 4x the normal rate of society,” he shared.
At Hives for Heroes, participants are given a new purpose… saving the declining bee population.
Jesse Puckett with the renewable energy company Enel said when construction started on the Roseland Solar Farm, they began looking for a beekeeping partner.
“We were looking for ways to increase biodiversity on our site. You know through our sustainability initiatives a lot of our work is supporting pollinator friendly habitats,” Puckett said.
0comments
That’s how they found Hives for Heroes, approaching them in 2024.
“One of the things that we were looking for is a partner that has really great community influence,” Puckett shared, “Hives for Heroes really checks a lot of those boxes for us so it’s a really great partner for us to have.”
The site now has a fully functioning hive that makes honey and pollinates the surrounding area.
“We have about 4×4 square foot, which is all we need! Which is wonderful cause then the bees take the rest of the land,” Jimenez said.
Jimenez says for many veterans and first responders the hands-on work has become a passion, adding that they’re now helping educate others.
“Other veterans and first responders, even schools, have come out to see the bees,” he shared, “we put them in the gear and they go through like inspections for example, and they’re educating the community and that’s a huge component of the outreach.”.
Hives for Heroes currently has bee hives at two Enel solar farms in Texas, one here in Central Texas and another near Bryan-College Station.
Both Jimenez and Puckett said they hope to grow that number over the next few years.
“We would love to see how our projects can continue to provide that land access, which is a barrier for many that are looking to get into beekeeping,” Puckett said.
Those interested can sign up here. No experience is needed as the organization provides all training for free.
Copyright 2026 KWTX. All rights reserved.
We want to hear what you think. Create a free account to share your thoughts. Read our community guidelines here.

source

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 16, Article number: 10336 (2026) Cite this article
1525 Accesses
Metrics details
Accurate photovoltaic (PV) power forecasting is essential for grid operation but remains difficult due to nonlinear multi-scale dynamics and seasonal distribution shifts. This work presents MKAN-iTransformer, a cascaded framework that integrates two existing components—the Multi-Scale Kolmogorov–Arnold Network (MKAN) for scale-aware temporal representation learning and iTransformer for variable-wise attention and inter-variable dependency modeling—under a 15-minute single-step setting. Experiments on a real-world 30 MW PV plant dataset from the Chinese State Grid Renewable Energy Generation Forecasting Competition use chronological splits within each season. MKAN-iTransformer achieves the best overall performance in spring, autumn, and winter. In spring, it reaches MSE=2.892, RMSE=1.701, MAE=0.864, and ({R^{2}}=0.947), improving over LSTM by 23.5%/12.5%/20.5% (MSE/RMSE/MAE). In autumn, it attains MSE=2.884, RMSE=1.698, MAE=0.774, and ({R^{2}}=0.962), reducing errors vs. iTransformer by 16.5%/8.7%/12.4%. In winter, it achieves MSE=1.721, RMSE=1.312, MAE=0.443, and ({R^{2}}=0.969), yielding 81.6%/57.1%/71.9% error reductions vs. Transformer. Ablation further confirms the complementarity between MKAN and iTransformer and shows that direct KAN integration can be unstable under winter shifts (KAN-iTransformer: MSE=7.082, ({R^{2}}=0.872)).
Amid escalating global climate change, transforming energy structures and accelerating renewable energy adoption have become shared priorities worldwide¹. The growing electricity demand drives increasing renewable power requirements², motivated by the carbon neutrality and eco-friendliness of renewable energy sources (RESs) compared to fossil fuels³. Empirical studies report negative associations between carbon emissions and renewable energy consumption, indicating emissions decrease as per-capita renewable energy use increases^4,5.
According to the International Energy Agency (IEA), renewables are expected to supply 42% of global electricity generation between 2023 and 2028, with solar and wind contributing 25%⁶. Photovoltaic (PV) generation, as representative green technology, has experienced rapid expansion. Global PV installed capacity continues growing with steadily rising power system share⁷. While large-scale PV integration brings substantial environmental benefits, it introduces new operational challenges⁸.
The primary challenge stems from PV output variability. PV generation is highly sensitive to meteorological conditions—irradiance, temperature, humidity, and wind speed—whose nonlinear and time-varying nature creates pronounced fluctuations and uncertainty⁹. This uncertainty complicates grid dispatch, increases storage and flexibility requirements, and affects electricity market operations^10,11. Therefore, accurate PV power forecasting is essential for secure and efficient power system operation¹² and supports downstream decision-making including operational management and demand response¹³.
These observations motivate a design that (i) captures multi-scale temporal dynamics of PV series and (ii) models cross-variable dependencies among meteorological inputs and historical power, while supporting transparent analysis. We develop MKAN-iTransformer, integrating Multi-Scale Kolmogorov-Arnold Networks (MKAN)¹⁴ with iTransformer¹⁵. MKAN provides multi-scale temporal representation learning with explicit functional structure, while iTransformer models inter-variable dependencies through variable-wise attention, together targeting robust and interpretable PV forecasting.
Contributions. Our main contributions are:
Cascaded forecasting framework. We develop MKAN-iTransformer, cascading multi-scale temporal representation learning with variable-wise attention for 15-minute single-step PV power prediction.
Season-wise chronological evaluation. Beyond overall test splits, we evaluate models within each season using chronological splits, making seasonal robustness and failure modes explicit.
Comprehensive KAN-enhanced baseline construction and evaluation. We systematically construct KAN/MKAN-augmented variants of recurrent and attention-based baseline architectures and establish unified benchmarking framework with consistent preprocessing, training, and evaluation protocols, enabling fair comparison and demonstrating the broad applicability of interpretable neural components in PV forecasting.
Interpretability analysis. We provide multi-scale time-frequency decomposition, learned KAN function inspection, and attention visualization for transparent explanations.
PV power forecasting has progressed from physics-driven and classical statistical models to modern machine learning and deep learning pipelines, largely driven by the need to handle nonlinearity, non-stationarity, and regime shifts.
Physical and statistical models. Early forecasting relied on physical simulation using meteorological inputs and device characteristics, which can be physically meaningful but often requires high-quality inputs and detailed plant specifications, limiting scalability in practice^9,12. Classical statistical models (e.g., ARMA/ARIMA and regression families) exploit temporal correlations and can perform well under relatively stable conditions; for instance, regression combined with numerical weather prediction has shown robust hourly forecasting¹⁶. However, abrupt ramps and distribution shifts common in PV generation challenge these assumptions and motivate more flexible nonlinear approaches.
Machine learning approaches. Conventional ML methods improved nonlinear mapping from weather variables to PV output, including linear regression and ensemble methods such as random forests and gradient boosting^17,18. Support vector regression has also been adopted for high-dimensional nonlinear forecasting^17,19. Despite progress, many ML pipelines rely on handcrafted features and can degrade under seasonal and weather-regime shifts, encouraging end-to-end deep architectures with better representation learning.
RNN-based models. LSTM and GRU variants have been widely used to capture temporal dependencies in PV forecasting^20,21,22. Performance gains have been reported via parallel structures, feature selection, CNN integration, and attention augmentation^20,23,24,25. Nevertheless, RNN-based components remain sequential and may become computational bottlenecks for long contexts, while their ability to explicitly model cross-variable interactions is often limited.
Hybrid CNN-RNN architectures. CNN-LSTM and related hybrids seek to combine local pattern extraction and temporal modeling²⁶, with variants replacing standard CNN blocks by temporal convolutional networks to improve receptive fields and parallelism²⁷. Attention mechanisms are frequently introduced for feature weighting and fusion; for example, dual-stream CNN-LSTM with self-attention has been reported to improve accuracy on PV datasets²⁸. However, these hybrids may still struggle to represent multi-scale behaviors spanning intra-hour variability to seasonal cycles, and they often treat heterogeneous meteorological variables as homogeneous inputs without an explicit variable-wise dependency mechanism.
Transformer-based architectures. Transformers have enabled stronger long-range dependency modeling for time series, with forecasting-oriented variants targeting efficiency and inductive biases. Informer reduces attention complexity via ProbSparse attention with (O(L log L)) behavior²⁹; Autoformer and FEDformer incorporate decomposition and frequency-aware mechanisms to better capture trend/seasonality^30,31. In PV-specific contexts, multi-scale and hybrid designs combine Transformers with CNNs/GRUs or decomposition modules^32,33,34, and domain-enhanced Transformers inject domain knowledge or nonlinear dependency modeling to improve robustness^35,36. The iTransformer introduces an inverted design that treats variables (rather than time steps) as tokens, enabling efficient variable-wise attention for cross-variable dependency modeling¹⁵, which is particularly relevant for PV forecasting where meteorological drivers and historical power jointly determine future output.
Multi-scale pattern recognition. PV generation exhibits multi-scale dynamics (diurnal cycles, intra-hour fluctuations, and weather-driven ramps), motivating multi-resolution modeling through decomposition, frequency-aware transformations, or multi-scale feature extraction. Interpretable deep learning pipelines have been proposed to disentangle multi-scale solar radiation variations while retaining predictive accuracy (e.g., reporting (R^2=0.97))³⁷. Yet, many approaches increase architectural complexity and do not always provide transparent, component-wise explanations that remain stable across operating regimes.
Interpretability requirements in energy systems. For energy applications, interpretability supports operational decision-making and stakeholder trust, but many deep models remain black boxes; moreover, attention weights alone do not guarantee faithful explanations. This motivates exploring model families with more explicit functional forms.
Kolmogorov–Arnold Networks (KAN) for interpretable learning. KANs parameterize multivariate mappings via sums of learned univariate functions, often implemented with spline-based learnable functions, offering a potentially more inspectable representation than standard MLP layers³⁸. Recent surveys summarize rapid development of KAN variants and applications (e.g., TKAN, Wav-KAN, DeepOKAN) and discuss their empirical strengths^38,39. Theoretical extensions such as KKANs further improve robustness and approximation behavior⁴⁰. For temporal data, KAN-based time series modeling has been explored, including general demonstrations and targeted work on bridging accuracy and interpretability in time series settings^41,42. KAN integration with dynamical systems has also been studied via KAN-ODEs⁴³. More recently, multi-scale KAN variants (MKAN) have been proposed to better capture mixed-frequency behaviors in temporal signals¹⁴. Despite these developments, systematic integration of KAN-style multi-scale representations with state-of-the-art variable-wise attention, and task-specific interpretability validation for PV forecasting, remains limited.
Many PV forecasting studies emphasize aggregate metrics, which can obscure failure modes under seasonal regime shifts. Seasonal variability changes irradiance, temperature, and daylight duration, making season-wise evaluation important for deployment^9,12. However, evaluation protocols and baselines are often inconsistent across model families, hindering fair comparison and limiting insights into robustness under regime transitions.
The above literature motivates four gaps addressed in this work:
Architectural integration gap: limited evidence on combining multi-scale temporal representations with explicit variable-wise attention for PV forecasting^14,15.
Interpretability integration gap: insufficient task-specific validation of interpretability when integrating KAN-style components with attention-based architectures^38,39.
Evaluation methodology gap: limited systematic assessment under seasonal regime shifts^9,12.
Benchmarking consistency gap: inconsistent protocols across model families impede fair comparison and understanding of when interpretable neural components help²¹.
Building on existing components^14,15, our MKAN-iTransformer focuses on principled integration of MKAN-style multi-scale representation learning with variable-wise attention, accompanied by systematic seasonal evaluation and interpretability-oriented analyses to clarify both strengths and limitations under different regimes.
The photovoltaic (PV) power forecasting task aims to predict the near-future output power of a PV plant based on historical multivariate time series observations. Let the historical observation sequence be
where (x_t in mathbb {R}^d) denotes the d-dimensional feature vector at time step t.
In this study, we consider single-step forecasting with a 15-minute horizon (sampling interval = 15 minutes). Therefore, the forecasting horizon is (h=1), and the prediction target is the PV power at the next time step:
The input variables include total solar irradiance, direct normal irradiance, global horizontal irradiance, air temperature, atmospheric pressure, relative humidity, and historical PV power. The target variable is the PV plant output power at the next 15-minute step.
The Multi-Scale Kolmogorov-Arnold Network (MKAN) module is designed to efficiently capture complex, multi-scale temporal dependencies in multivariate time series forecasting. The overall structure is illustrated in Fig. 1 and consists of the following key components.
Overall architecture of the Multi-Scale Kolmogorov-Arnold Network (MKAN) module. The left part shows the hierarchical residual structure with stacked TimeKAN blocks, each extracting features at different scales through multi-scale patching (MSP) modules. The right part details the patching, encoding, KAN-based transformation, decoding, and unpatching process within each MSP block. Cumulative addition and subtraction operations are used to aggregate both local and global temporal features.
Multi-scale patching: Given an input sequence (X in mathbb {R}^{T times d}), we divide it into S sets of patches at different temporal scales, where the s-th scale consists of (N_s) patches of length (l_s):
Patch encoder: Each patch is mapped to a latent embedding via a learnable encoder:
where (operatorname {Enc}_s) denotes the patch encoder for scale s.
KAN-based Transformation: Each scale has a dedicated Kolmogorov-Arnold Network (KAN) block to transform the encoded patch embedding:
where (operatorname {KAN}_s) is the KAN subnetwork for the s-th scale.
Patch decoder: The transformed embeddings are decoded back to the temporal domain:
Feature aggregation: The reconstructed patches are reassembled to form multi-scale feature maps, which are then aggregated (e.g., by summation or concatenation) to obtain the final sequence representation:
where (operatorname {Agg}) denotes the aggregation operation across scales.
Forecasting head: The aggregated features are passed to a forecasting head to generate the final prediction:
The overall output of the MKAN module can be summarized as a weighted sum of KAN transformations across all scales:
where (phi _{s,n}(cdot )) represents the output of the KAN subnetwork for the n-th patch at scale s, (alpha _{s,n}) are learnable weights, and b is a bias term.
A major advantage of the MKAN module is its interpretability. Each KAN block is inherently symbolic and can be visualized or analyzed, allowing for direct inspection of the learned temporal features at each scale. In summary, the MKAN module integrates multi-scale patching with expressive KAN transformations, providing a transparent and effective solution for multivariate time series forecasting.
The iTransformer module is designed to efficiently model multivariate time series forecasting by leveraging an inverted Transformer architecture. The overall structure is illustrated in Fig. 2 and consists of the following key components.
Overall architecture of the iTransformer module. The framework consists of independent variable-wise embedding, temporal layer normalization, multivariate self-attention, feed-forward transformation, and aggregation for final forecasting. The left and right parts of the figure detail the embedding and feed-forward processes, respectively.
Variable-wise embedding: Given a multivariate input sequence (X in mathbb {R}^{T times N}), where T is the sequence length and N is the number of variables, each variable’s time series (X_{:,n}) is independently embedded into a latent representation:
where (operatorname {Embedding}) is a learnable mapping from (mathbb {R}^T) to (mathbb {R}^d).
Temporal layer normalization: Each variable embedding is normalized along the temporal dimension to reduce scale and distribution discrepancies:
where (mu _n) and (sigma _n) are the mean and standard deviation of the n-th variable embedding.
Multivariate self-attention: All variable embeddings are jointly processed by a self-attention mechanism to capture inter-variable dependencies:
where Q, K, V are linear projections of the variable embeddings. The detailed structure of the multivariate self-attention mechanism is shown in Fig. 3.
Detailed structure of the multivariate self-attention mechanism in the iTransformer module. The input is first projected to Q, K, and V, then split into multiple heads for independent attention computation. The results are merged and projected to form the final output.
Feed-forward network: Each variable embedding is independently transformed by a shared feed-forward network to extract nonlinear features:
where (operatorname {FFN}) denotes a two-layer MLP with activation and dropout.
Stacked blocks and aggregation: The above operations are stacked for L layers, and the final output embeddings are aggregated for forecasting:
where (operatorname {TrmBlock}) denotes one iTransformer block, and (operatorname {Projection}) maps the final embedding to the prediction space.
The overall output of the iTransformer module can be summarized as:
where (operatorname {Head}) is typically a linear layer for regression or forecasting.
A major advantage of the iTransformer module is its variable-centric design. By treating each variable’s time series as an independent token, the model can explicitly capture inter-variable correlations and global temporal patterns, while maintaining efficient parallel computation and interpretability of learned representations. In summary, the iTransformer module integrates variable-wise embedding, normalization, and attention-based transformation, providing a simple yet powerful backbone for multivariate time series forecasting.
The hybrid architecture adopts a cascaded design, where the Multi-Scale Kolmogorov-Arnold Network (MKAN) module first extracts multi-scale temporal features from the input sequence, and the resulting representations are subsequently processed by the iTransformer module to model inter-variable dependencies. The overall structure is illustrated in Fig. 4.
Overall architecture of the cascaded MKAN-iTransformer framework. The pipeline consists of sequential MKAN and iTransformer modules, followed by a forecasting head. The left part details the multi-scale patching and KAN transformation, while the right part illustrates variable-wise attention and prediction.
Multi-scale feature extraction (MKAN): Given an input sequence (X in mathbb {R}^{T times N}), the MKAN module extracts multi-scale temporal features:
where (Z_{text {MKAN}}) encodes rich temporal dependencies across different resolutions.
Inter-variable modeling (iTransformer): The multi-scale features (Z_{text {MKAN}}) are fed into the iTransformer module, which captures global dependencies among variables via self-attention mechanisms:
where (Z_{text {iTrm}}) denotes the variable-attentive feature representation.
Forecasting head: The final representation is passed to a forecasting head to generate the prediction:
This cascaded hybrid design enables the model to:
Efficiently extract multi-scale temporal patterns using the MKAN module, which models complex dynamics at various time resolutions.
Explicitly capture inter-variable relationships through the iTransformer, which leverages attention to integrate information across variables.
Produce robust and interpretable representations for accurate multivariate time series forecasting.
In summary, the cascaded MKAN-iTransformer architecture unifies multi-scale temporal feature extraction and variable-wise attention modeling, forming a transparent and powerful backbone for multivariate time series forecasting.
In this study, real-world operational data from a 30 MW photovoltaic (PV) power plant are utilized for experimental evaluation. The dataset contains records from 2019 and 2020, with a sampling interval of 15 minutes. The input features include total solar irradiance, direct normal irradiance, global horizontal irradiance, air temperature, atmospheric pressure, and relative humidity. The target variable is the output power of the PV power plant. Details are shown in Table 1.
The quality of the dataset has a decisive impact on the accuracy of forecasting models. Therefore, it is particularly important to pay attention to missing value handling and dataset partitioning during the process of model optimization. To ensure the overall trend and consistency of the data, this study first employs linear interpolation to impute missing values during the data preprocessing stage. For outliers in each column, reasonable value ranges are defined based on actual physical meanings. Values exceeding these ranges are clipped to the valid interval, thereby improving the reliability of the data and the prediction accuracy of the model.
Due to the fact that the operational intensity of photovoltaic systems is almost negligible during nighttime, the dataset contains sparse and uninformative data points for these periods. Such sparsity is detrimental to the performance of forecasting models. To address this issue, all nighttime data points were excluded from the dataset in this study. Specifically, only data collected between 6:00 AM and 8:00 PM were retained for subsequent experiments.For the 15-minute single-step setting, we align inputs and targets by shifting the PV power series by one step: the target at time t is the PV power at (t+1). This alignment is performed after nighttime filtering, and no future information is included in the model inputs.
A total of 70,177 sampling points were collected from two years of photovoltaic data. The data were divided into four seasons according to the following scheme: spring (March to May), summer (June to August), autumn (September to November), and winter (December to February). The number of sampling points for each season was 17,666, 17,378, 17,467, and 17,666, respectively.
To explore the relationships between meteorological and operational features and photovoltaic (PV) output power, this study employs the Pearson Correlation Coefficient for all numerical variables. The Pearson correlation coefficient measures the degree of linear correlation between two variables, with possible values in the interval ([-1,1]), where a value closer to 1 or (-1) indicates a stronger correlation. A positive value indicates a positive correlation, while a negative value indicates a negative correlation. Note that the correlation analysis is conducted for interpretability and exploratory understanding, rather than for feature selection. In particular, we retain all physically meaningful variables to support the subsequent variable-wise attention visualization of the iTransformer and to avoid excluding variables that may contribute through nonlinear interactions.
The calculation formula for the Pearson correlation coefficient is as follows:
where (x_i) and (y_i) denote the i-th observations of the two variables, (bar{x}) and (bar{y}) are their respective means, and n is the total number of samples.
The correlation among features is visualized in the form of a heatmap, as shown in Fig. 5. Furthermore, the Pearson correlation coefficients between the main meteorological features and PV output power are listed in Table 2.
Heatmap of Pearson correlation coefficients among main features.
As shown in Fig. 5 and Table 2, PV output power (Power, MW) has the strongest correlation with total solar irradiance (Total solar irradiance, W/m(^2)), with a coefficient as high as 0.95. It also shows strong positive correlations with direct normal irradiance (Direct normal irradiance, W/m(^2)) and global horizontal irradiance (Global horizontal irradiance, W/m(^2)), with coefficients of 0.89 and 0.64, respectively. This indicates that irradiance is the dominant factor affecting PV output power.
Air temperature ((^circ)C) has a correlation coefficient of 0.26 with output power, indicating a weak positive correlation. Relative humidity (%) shows a negative correlation with output power, with a coefficient of (-0.35). Atmospheric pressure (hPa) exhibits a very low correlation with PV output power, suggesting a limited linear association. Overall, irradiance-related features are the primary factors influencing PV output power, while temperature, humidity, and pressure provide complementary meteorological information.
Final input features. In the forecasting experiments, the model inputs include total solar irradiance, direct normal irradiance, global horizontal irradiance, air temperature, atmospheric pressure, relative humidity, and historical PV power, while the prediction target is the PV plant output power at the next 15-minute step.
This subsection describes the compared models and the unified hyperparameter tuning protocol used to ensure fair and reproducible evaluation.
Compared models. We evaluate multiple forecasting backbones and their KAN/MKAN-augmented variants for 15-minute single-step PV power forecasting. KAN and MKAN are adopted from prior work; we implement their integrations with different backbones to form the compared variants. Specifically, we consider LSTM/GRU/BiLSTM/Transformer/xLSTM/iTransformer and their corresponding KAN- and MKAN-augmented versions (i.e., KAN-LSTM and MKAN-LSTM; KAN-GRU and MKAN-GRU; KAN-BiLSTM and MKAN-BiLSTM; KAN-Transformer and MKAN-Transformer; KAN-xLSTM and MKAN-xLSTM; KAN-iTransformer and MKAN-iTransformer). All models are trained and evaluated under the same input–output setting.
Chronological split. To avoid look-ahead bias in time-series forecasting, we split the data in chronological order into a training set (80%), a validation set (10%), and a test set (10%). Specifically, the earliest 80% of samples are used for training, the subsequent 10% for validation, and the latest 10% for testing. The same temporal rule is applied within each seasonal subset.
Grid search protocol. Hyperparameters are tuned on the validation set using a grid search with the following candidate values: learning rate in ({1times 10^{-2}, 5times 10^{-3}, 1times 10^{-3}, 5times 10^{-4}}), hidden dimension in ({32, 64, 128}), number of skip connections in ({1, 2, 3}), number of attention heads in ({2, 4, 8}), and convolution kernel size in ({3, 5, 7}). This yields (4times 3times 3times 3times 3 = 324) configurations.
For model components where a hyperparameter is not applicable (e.g., attention heads for purely recurrent architectures), we keep that component at its default setting while tuning the remaining applicable parameters. The same tuning criterion (minimum validation loss) and training budget are applied to all models.
Training and selection. Each configuration is trained for up to 100 epochs with early stopping based on the validation loss (patience = 10), and the checkpoint with the best validation loss is selected. The best hyperparameter setting is chosen according to the validation loss. Using the selected hyperparameters, we retrain the model on the union of the training and validation sets and report the final performance on the held-out test set. All experiments are conducted with a fixed random seed (seed = 42) to reduce randomness.
All models are implemented in PyTorch and trained using the same pipeline to ensure a fair comparison. The input features are standardized using statistics computed on the training split only, and the same transformation is applied to the validation and test splits. The PV power target is kept in its original scale (i.e., no target normalization is applied).
We optimize all models using the Adam optimizer and minimize the mean squared error (MSE) on the training set. The initial learning rate and other hyperparameters are selected via the validation-based grid search described above. We use mini-batch training with a batch size of 64. To improve training stability, gradient clipping is applied with a maximum norm of 1.0. Early stopping is performed based on the validation loss with a patience of 10 epochs, and the checkpoint with the lowest validation loss is selected.
After hyperparameter selection, each model is retrained on the combined training and validation sets using the selected configuration, and the final performance is reported on the held-out test set using MSE, RMSE, MAE, and (R^2). All experiments are conducted with a fixed random seed (seed = 42) to reduce randomness.
To comprehensively evaluate the prediction performance of the proposed MKAN-iTransformer and baseline models, four commonly used regression metrics are adopted: Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination ((R^2)). The definitions are as follows:
Mean Squared Error (MSE):
Root Mean Squared Error (RMSE):
Mean Absolute Error (MAE):
Coefficient of determination ((R^2)):
where (bar{y}) is the mean of the true values.
A lower value of MSE, RMSE, and MAE indicates better model performance, while a higher (R^2) value (closer to 1) implies a better fit between predictions and actual values.
In this section, we present and analyze the experimental results of the proposed MKAN-iTransformer model and various baseline methods on photovoltaic power forecasting tasks. The experiments are conducted under different seasonal. The performance of all models is evaluated using the metrics introduced previously.
To systematically evaluate the impact of seasonal variations on model performance, we adopted the conventional monthly division method to classify the dataset into four seasons: spring (March–May), summer (June–August), autumn (September–November), and winter (December–February). This classification enables a more detailed analysis of the predictive capabilities of MKAN-iTransformer and baseline models under different seasonal conditions.
Typical daily PV power curves for each month in 2019 and 2020.
As shown in Fig. 6, the typical daily power curves for each month in 2019 and 2020 exhibit significant seasonal variations. The power output is higher in spring and summer due to abundant sunlight, while it is relatively lower in autumn and winter. These seasonal differences provide a solid foundation for the subsequent model performance analysis based on seasonal classification.
To validate the effectiveness of MKAN-iTransformer, we conducted a detailed analysis of model prediction results and error distributions across different seasonal conditions. This section presents a comparative evaluation of MKAN-iTransformer and baseline models, highlighting both the accuracy and robustness of the proposed approach.
Spring: single-day prediction curves and prediction error distribution.
Summer: single-day prediction curves and prediction error distribution.
Autumn: single-day prediction curves and prediction error distribution.
Winter: single-day prediction curves and prediction error distribution.
Figures 7, 8, 9 and 10 present typical-day forecasting results for spring, summer, autumn, and winter, respectively. For each season, the upper subfigure compares the predicted PV output power with the ground-truth measurements (black curve) over the daytime period (6:00–20:00 at 15-minute intervals), while the lower subfigure summarizes the corresponding prediction error distribution of each model. This season-wise “curve fitting + error distribution” layout allows an intuitive assessment of both temporal tracking ability (shape, peak timing, and ramping behavior) and statistical error characteristics (bias, dispersion, and tail behavior).
In the spring case (Fig. 7), most models can capture the overall diurnal pattern, but noticeable deviations appear around rapid ramping segments and local peaks. The proposed MKAN-iTransformer shows a closer alignment with the ground truth during the main rising stage and peak region, and its error histogram is more concentrated around zero, suggesting reduced dispersion and fewer large-magnitude errors.
For summer (Fig. 8), PV output typically exhibits a smoother and higher plateau under stronger irradiance conditions, making the dominant daily trend easier to learn. Accordingly, multiple models achieve relatively good tracking performance. Nevertheless, differences remain in reproducing sharp changes (e.g., abrupt drops and recoveries), where MKAN-iTransformer tends to maintain smaller deviations. The error distribution in summer is comparatively narrower for several models, indicating that the forecasting task is less challenging than in transitional or winter conditions.
In autumn (Fig. 9), the ground-truth curve shows more frequent fluctuations and irregular ramps, likely due to increased variability in meteorological conditions. Some baseline models display either lagged responses or oversmoothing, leading to larger deviations during abrupt changes. MKAN-iTransformer provides more stable tracking across multiple fluctuation segments, and its error distribution shows reduced spread relative to many baselines, implying improved generalization to more volatile patterns.
Winter results (Fig. 10) are the most challenging, as reflected by larger mismatches in several baselines and a visibly broader error spread in the histogram. The seasonal difficulty may be attributed to lower sun angles, shorter effective generation windows, and more frequent rapid variations (e.g., due to clouds and atmospheric conditions), which amplify both bias and variance in predictions. In contrast, MKAN-iTransformer remains closely aligned with the ground truth for most time intervals, and the error distribution remains comparatively concentrated, indicating stronger robustness under adverse seasonal conditions.
Overall, across all four seasons, MKAN-iTransformer consistently achieves closer curve fitting and more compact error distributions, demonstrating improved accuracy and robustness. These observations are consistent with the quantitative seasonal metrics reported in Table 3, where MKAN-iTransformer achieves competitive or best performance on MSE/RMSE/MAE and high (R^2) in multiple seasons.
Using MKAN-iTransformer as the main benchmark, we compare it with representative baselines (LSTM, GRU, BiLSTM, Transformer, xLSTM, and iTransformer) as well as KAN/MKAN-augmented variants on seasonal datasets. The quantitative results in Table 3 indicate that MKAN-iTransformer achieves the most consistent and competitive performance across seasons. In particular, it attains the best overall results in spring, autumn, and winter (covering MSE, RMSE, MAE, and (R^2)), while in summer it delivers the lowest MSE/RMSE and remains highly competitive in (R^2), although the best MAE is achieved by KAN-GRU.
In spring, MKAN-iTransformer achieves the best performance across all four metrics, with MSE = 2.892, RMSE = 1.701, MAE = 0.864, and (R^2) = 0.947. Compared with LSTM, it reduces MSE/RMSE/MAE by 23.5%, 12.5%, and 20.5%, respectively, and improves (R^2) by 1.7%. Relative to GRU, MKAN-iTransformer reduces MSE by 13.7%, RMSE by 7.1%, and MAE by 13.4%, while increasing (R^2) by 0.9%. Against Transformer, the reductions are 7.4% (MSE), 3.7% (RMSE), and 11.5% (MAE), with a 0.4% gain in (R^2). These improvements demonstrate that MKAN-iTransformer better captures springtime ramping and peak behaviors, yielding both lower average error and improved goodness-of-fit.
Summer exhibits different characteristics: MKAN-iTransformer achieves the lowest MSE (3.962) and RMSE (1.991) among all compared models, while the best MAE is obtained by KAN-GRU (0.951), and the highest (R^2) is achieved by xLSTM (0.924). Compared with LSTM, MKAN-iTransformer decreases MSE and RMSE by 3.3% and 1.6%, and slightly increases (R^2) (0.921 to 0.923). Compared with iTransformer, it yields a clear reduction in MSE (9.5%) and RMSE (4.8%) and improves (R^2) from 0.915 to 0.923. Although its MAE is not the best in summer, the advantage in MSE/RMSE suggests MKAN-iTransformer is particularly effective at suppressing larger deviations (which are weighted more heavily by MSE), while some models (e.g., KAN-GRU) achieve smaller absolute errors on average.
In autumn, MKAN-iTransformer again provides the best results across all metrics (MSE = 2.884, RMSE = 1.698, MAE = 0.774, (R^2) = 0.962). Compared with LSTM, it reduces MSE/RMSE/MAE by 24.9%, 13.4%, and 27.4%, respectively, and improves (R^2) by 1.4%. Relative to GRU, it reduces MSE by 9.4%, RMSE by 4.8%, and MAE by 11.6%, with (R^2) increasing from 0.958 to 0.962. Compared with iTransformer, MKAN-iTransformer reduces MSE by 16.5%, RMSE by 8.7%, and MAE by 12.4%, while improving (R^2) from 0.954 to 0.962. These results indicate stronger adaptability to autumn’s higher variability and more frequent fluctuations.
Winter is the most challenging season for many baselines, yet MKAN-iTransformer achieves the strongest overall performance with MSE = 1.721, RMSE = 1.312, MAE = 0.443, and (R^2) = 0.969. Compared with LSTM, it reduces MSE/RMSE/MAE by 71.4%, 46.5%, and 66.6%, respectively, and improves (R^2) from 0.891 to 0.969 (an 8.8% relative increase). Against Transformer, the reductions are 81.6% (MSE), 57.1% (RMSE), and 71.9% (MAE), with (R^2) increasing from 0.831 to 0.969. Compared with iTransformer, MKAN-iTransformer remains slightly better in error-based metrics (e.g., MSE from 1.730 to 1.721 and RMSE from 1.315 to 1.312) while maintaining the same (R^2). Overall, these results demonstrate that MKAN-iTransformer offers strong robustness under winter conditions, substantially reducing both average errors and large-error events relative to most baselines.
This work focuses on evaluating the effectiveness of combining an iTransformer backbone with KAN-based modules. Note that KAN and MKAN are borrowed from prior work and are not proposed in this paper; our goal is to investigate whether integrating these modules with iTransformer yields complementary gains and improved robustness across seasonal distributions.
Model variants. We compare four variants: (1) iTransformer, the backbone baseline; (2) KAN-iTransformer, which integrates a KAN-based (ekan) module into iTransformer; (3) MKAN-iTransformer, which combines MKAN with iTransformer (our main combination model); and (4) MKAN, the standalone MKAN model without iTransformer, included to distinguish the effect of MKAN alone from the fusion setting. All variants are trained and evaluated under the same experimental protocol.
Metrics. We report MSE, RMSE, and MAE (lower is better) as well as (varvec{R^2}) (higher is better). To examine distribution shifts, results are presented for Spring, Summer, Autumn, and Winter.
Results and discussion. As shown in Table 4, MKAN-iTransformer delivers the most consistent improvements across seasons. In Spring, it achieves the best results on all metrics, indicating clear complementarity between MKAN and iTransformer. In Autumn, MKAN-iTransformer again obtains the best overall performance, slightly outperforming KAN-iTransformer, suggesting that the multi-scale design provides additional benefit beyond directly integrating KAN.
In Summer, MKAN-iTransformer yields the lowest MSE/RMSE and the highest (R^2), while iTransformer attains the lowest MAE. This indicates a trade-off between reducing larger errors (more reflected by squared-error metrics) and minimizing average absolute deviation; nevertheless, the improved RMSE and (R^2) suggest a better overall fit for MKAN-iTransformer.
In Winter, iTransformer and MKAN-iTransformer are nearly identical, implying that the iTransformer backbone already captures the dominant winter dynamics and that MKAN integration does not introduce degradation. By contrast, KAN-iTransformer shows a pronounced performance drop in winter (MSE=7.082, (R^2)=0.872), indicating that this integration may be more sensitive to seasonal distribution shifts. Overall, these results support that MKAN-iTransformer is a robust and effective combination, whereas the gains from KAN-iTransformer are less stable across seasons.
To explain the seasonal performance differences observed in the previous sections, we conduct an interpretability analysis of MKAN from three complementary perspectives. First, we decompose the PV power signal into hierarchical temporal components to isolate fast ramps, intermediate variations, and slow diurnal trends, and validate the separation in both time and frequency domains. Second, we inspect the learned KAN edge functions to understand how MKAN adapts its nonlinear transformations across seasons. Third, we visualize the inverted attention mechanism over features to quantify seasonal changes in feature importance, attention dispersion, and cross-feature interaction pathways.
Together, these analyses form a consistent evidence chain from signal dynamics (multi-scale decomposition), to nonlinear representation (KAN activations), and finally to decision routing (feature-wise attention), clarifying why the model behaves differently under distinct seasonal atmospheric regimes.
To capture PV dynamics from fast cloud-induced ramps to slow diurnal trends, the MKAN module decomposes the 15-min PV power series into three additive temporal components using hierarchical moving-average (MA) operators and residual (difference) bands. This formulation yields a physically consistent separation of high-, mid-, and low-frequency behaviors while preserving approximate additivity.
Let P(t) denote the normalized PV power at 15-min resolution and let (textrm{MA}_m(cdot )) be an m-step moving average (centered window for analysis/visualization). We define:
Thus,
where (epsilon (t)) mainly captures boundary effects and minor mismatch.
Figure 11 illustrates the multi-scale decomposition of PV power on a representative summer day. The decomposition separates the observed signal into three time-scale components, which helps interpret variability sources and motivates using scale-aware features in forecasting.
High-frequency (45 min and below): rapid ramps and short-term fluctuations dominated by transient clouds and local turbulence, critical for short-horizon forecasting.
Medium-frequency (90–180 min): intra-day variability related to evolving weather regimes and smooth changes in solar geometry.
Low-frequency (180 min trend): slowly varying baseline reflecting the dominant diurnal envelope and seasonal irradiance level.
Multi-scale temporal decomposition of PV power on a representative summer day. From top to bottom, the panels show the original signal and its high-, medium-, and low-frequency components. The low-frequency term captures the smooth diurnal envelope, the medium-frequency term reflects intra-day regime changes, and the high-frequency term highlights fast fluctuations.
To validate that the proposed multi-scale decomposition indeed separates variability across time scales, we conduct a frequency-domain check using Welch’s power spectral density (PSD). Figure 12 reports the PSD characteristics of the decomposed components: the high-frequency residual (P_{text {high}}), the medium-frequency component (P_{text {mid}}), and the low-frequency trend (P_{text {low}}) for a representative summer day.
We partition the frequency axis into three bands (in cycles/hour) to summarize spectral energy:
Low-frequency band: (f < 0.1) (dominant diurnal/slow envelope and baseline variations).
Mid-frequency band: (0.1 le f le 0.5) (intra-day variability and regime transitions).
High-frequency band: (f > 0.5) (fast ramps and short-term fluctuations).
For each component, the band energy percentages are computed by integrating its PSD over the corresponding frequency band and normalizing by the component’s total spectral energy:
where
denotes the Welch PSD estimate and
is one of the three bands above.
As shown in Fig. 12, (P_{text {high}}) allocates a larger portion of energy to higher frequencies, while (P_{text {low}}) concentrates energy in the low-frequency region consistent with the diurnal envelope. The medium-scale component (P_{text {mid}}) mainly captures intermediate-band energy, supporting the intended multi-scale separation.
Frequency-domain validation (summer). The left column shows Welch PSD for (P_{text {high}}), (P_{text {mid}}), and (P_{text {low}}). The top-right panel compares PSD curves across scales, and the bottom-right panel summarizes the energy distribution over the predefined low/mid/high frequency bands.
Overall, this hierarchical MA residual decomposition provides interpretable temporal bands and supports MKAN’s multi-branch design, reducing interference between fast ramps and slow trends. The seasonal consistency of this separation is further confirmed in Table 5.
KAN replaces fixed activation functions (e.g., ReLU, GELU) with learnable univariate edge functions, making nonlinear transformations explicit and interpreable. We analyze learned activation patterns and relate their shapes to PV forecasting behavior across seasonal regimes.
For an input feature (x_i) and output node (y_j), KAN learns an edge function (phi _{i,j}(cdot )) using cubic B-splines:
where (B_k(x)) are spline basis functions, (c_{i,j,k}) are learnable coefficients, and K denotes the number of spline control points. A KAN layer aggregates edge functions as:
This formulation allows each connection to learn a data-driven nonlinear mapping tailored to a specific input-output relation.
Figure 13 illustrates representative learned KAN activations and their differences from standard fixed activations. In PV forecasting, asymmetric nonlinear responses are useful: suppressing low-power noise (e.g., dawn/dusk or heavy haze) while preserving sensitivity during normal operating conditions.
Comprehensive analysis of KAN activation functions. The figure compares fixed activations with representative learned KAN activations and highlights how learnable nonlinearities adapt to different data regimes.
Figure 14 shows season-specific learned activations, indicating that KAN adapts its nonlinearity to seasonal PV dynamics.
Seasonal adaptation of learned KAN activation functions. Each panel shows a representative learned activation from seasonal data (colored) compared with a fixed baseline (gray). Shaded regions indicate deviation, highlighting season-specific nonlinear adaptation.
We quantify seasonal differences using three metrics over a fixed input range:
Table 6 indicates stronger nonlinear adaptation in more challenging regimes, supporting KAN interpretability: learned activation shapes reflect seasonal PV generation characteristics.
MKAN adopts an iTransformer-style inverted attention mechanism operating over the feature dimension, enabling dynamic feature-to-feature interaction modeling. We visualize seasonal feature importance, attention distributions, and cross-feature attention pathways to interpret how meteorological variables contribute under different atmospheric conditions.
Figure 15 presents normalized feature importance by season. Table 7 reports the corresponding scores (normalized to the maximum within each season), revealing clear seasonal reweighting between irradiance-driven and atmosphere-driven predictors.
Seasonal comparison of feature importance scores. Bars show normalized importance of each meteorological feature within a season.
Across seasons, DNI dominates in spring and winter, while GHI becomes most important in summer, reflecting stronger scattering/cloud effects. Autumn shifts toward atmospheric pressure and historical power, suggesting increased reliance on synoptic conditions and temporal persistence during transitional weather.
Winter vs. summer shift: Compared with summer, winter assigns substantially higher importance to RH (+0.5742) and DNI (+0.4962), and also increases reliance on historical power (+0.3096). In contrast, GHI becomes less dominant in winter (–0.1831), consistent with reduced diffuse-driven regimes and stronger sensitivity to beam irradiance availability.
Figure 16 shows attention weight distributions across features and seasons. Wider distributions indicate more frequent reallocation of attention, typically associated with more volatile atmospheric conditions.
Seasonal attention weight distributions across features. F1: Total solar irradiance, F2: Direct normal irradiance, F3: Global horizontal irradiance, F4: Air temperature, F5: Atmospheric pressure, F6: Relative humidity, F7: Power.
We summarize attention dispersion using entropy computed from mean attention weights:
where (bar{w}_i) is the mean attention weight of feature i. Table 8 reports the attention entropy and the seasonal prediction performance (RMSE in MW). Higher entropy indicates more distributed attention (i.e., no single dominant feature), reflecting more frequent reallocation of attention across variables under volatile atmospheric conditions.
Figure 17 visualizes seasonal cross-feature attention matrices. To highlight dominant interaction pathways, Table 9 lists the top-3 attention pairs (query (rightarrow) key) per season.
Seasonal cross-feature attention matrices. Rows are query features, columns are key features. F1: Total solar irradiance, F2: Direct normal irradiance, F3: Global horizontal irradiance, F4: Air temperature, F5: Atmospheric pressure, F6: Relative humidity, F7: Power.
These pathways are physically plausible: summer emphasizes humidity–irradiance coupling (cloud formation and scattering), while winter concentrates multiple queries onto DNI, indicating that beam irradiance penetration becomes a key bottleneck signal under haze/fog conditions.
We further quantify attention matrix structure using diagonal dominance:
and interaction diversity:
Table 10 confirms a seasonal shift between distributed attention (higher I, lower D) and focused attention (higher D, lower I), consistent with changes in atmospheric conditions and feature reliability.
This study has several limitations that should be acknowledged when interpreting the results.
Single-site evaluation. All experiments are conducted on data from a single PV plant. While the seasonal split provides a meaningful within-site distribution-shift test, the cross-site generalization of MKAN-iTransformer (e.g., different climates, terrains, PV technologies, and sensor configurations) is not verified here.
Daytime-only forecasting protocol. Nighttime samples are excluded (06:00–20:00) because PV generation is near-zero and the series becomes sparse and less informative for learning daytime dynamics. This choice improves training stability and focuses the evaluation on operationally relevant generation periods, but it limits applicability to round-the-clock settings. In particular, behavior during dawn/dusk transitions and full-day forecasting is not evaluated.
Dataset size and coverage. The dataset covers two years and yields a moderate number of samples after filtering and seasonal partitioning. Although sufficient for 15-minute single-step forecasting, larger multi-year and multi-site datasets may expose additional failure modes, especially rare extreme-weather ramps.
Lack of uncertainty quantification. This work reports point forecasting metrics (MSE/RMSE/MAE and ({R^{2}})) only. For grid operation and risk-aware scheduling, probabilistic forecasts (e.g., prediction intervals or quantiles) and calibration analyses are often required. Uncertainty quantification is not addressed in this paper.
These limitations motivate future work on cross-site evaluation, round-the-clock and multi-horizon forecasting protocols, and probabilistic forecasting with calibrated uncertainty estimates.
This paper studies robust and interpretable PV power forecasting under seasonal regime shifts and proposes MKAN-iTransformer, a cascaded hybrid framework that combines MKAN-based multi-scale temporal representation learning with iTransformer-style variable-wise attention for cross-variable dependency modeling. The model is evaluated under a unified protocol for 15-minute single-step forecasting with consistent preprocessing, hyperparameter tuning, and chronological splits within each seasonal subset.
Seasonal accuracy and robustness. Season-wise benchmarking (Table 3) shows that MKAN-iTransformer achieves consistent and competitive performance across all four seasons. It delivers the best overall results in spring, autumn, and winter across MSE/RMSE/MAE and ({R^{2}}), and remains highly competitive in summer with the lowest squared-error metrics. The typical-day prediction curves and error histograms further support these findings by showing closer tracking during ramps and peaks and more concentrated error distributions, indicating fewer large-deviation events under seasonal variability.
Component contribution validated by ablation. The ablation study (Table 4) isolates the effects of MKAN and iTransformer and confirms that their combination is beneficial. Comparing iTransformer, MKAN, and MKAN-iTransformer demonstrates that neither multi-scale temporal modeling nor variable-wise dependency modeling alone fully explains the observed improvements; rather, the gains arise from their complementarity. In addition, the inclusion of KAN-iTransformer reveals that not all KAN-style integrations are equally stable: KAN-iTransformer exhibits a pronounced degradation in winter, suggesting sensitivity to seasonal distribution shifts, whereas MKAN-iTransformer remains robust.
Interpretability evidence. Beyond performance, we provide a coherent interpretability analysis from three perspectives: (i) a multi-scale temporal decomposition aligned with MKAN branches and validated in the frequency domain, clarifying how fast ramps, intermediate variations, and slow diurnal trends are separated; (ii) inspection and quantification of learned KAN univariate edge/activation functions, showing season-dependent nonlinear adaptations; and (iii) feature-wise attention visualization, demonstrating seasonal reweighting of meteorological drivers and physically plausible cross-feature interaction pathways.
Implications. Overall, MKAN-iTransformer offers an effective balance among accuracy, seasonal robustness, and model transparency for short-horizon PV forecasting. The results indicate that coupling scale-aware temporal feature extraction with explicit inter-variable modeling is a practical strategy to mitigate seasonal degradation commonly observed in baseline architectures.
Future directions. Future work will extend the evaluation to multi-site datasets and round-the-clock settings, generalize the framework to multi-horizon forecasting, and incorporate uncertainty quantification to produce calibrated prediction intervals suitable for risk-aware operational decision-making.
The dataset used in this study is sourced from the State Grid Corporation of China New Energy Power Generation Forecasting Competition (^{?}). Comprehensive experiments were conducted on this authoritative dataset to verify the superior performance of the proposed model under different seasonal conditions. The dataset is publicly available and can be accessed at the following link: https://www.nature.com/articles/s41597-022-01696-6#citeas.
Intergovernmental Panel on Climate Change. Climate Change: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. In Masson-Delmotte et al. (eds), Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA (2021). https://doi.org/10.1017/9781009157896
Ahmad, E. et al. The influence of grid connectivity, electricity pricing, policy-driven power incentives, and carbon emissions on renewable energy adoption: exploring key factors. Renew. Energy 232, 121108 (2024).
Article  CAS  Google Scholar 
Ullah, S. & Lin, B. Green energy dynamics: analyzing the environmental impacts of renewable, hydro, and nuclear energy consumption in pakistan. Renew. Energy 232, 121025 (2024).
Article  Google Scholar 
Perone, G. The relationship between renewable energy production and co2 emissions in 27 oecd countries: a panel cointegration and granger non-causality approach. J. Clean. Prod. 434, 139655 (2024).
Article  CAS  Google Scholar 
Zhao, C., Wang, J., Dong, K. & Wang, K. Is renewable energy technology innovation an excellent strategy for reducing climate risk? the case of china. Renew. Energy 223, 120042 (2024).
Article  Google Scholar 
Dechezleprêtre, A. et al. A comprehensive overview of the renewable energy industrial ecosystem. In Documents de Travail de l’OCDE sur la Science, la Technologie et l’industrie (2024).
Renewable energy statistics 2023, International Renewable Energy Agency, Abu Dhabi (2023). https://www.irena.org/-/media/Files/IRENA/Agency/Publication/2023/Jul/IRENA_Renewable_energy_statistics_2023.pdf
Agoua, X. G., Girard, R. & Kariniotakis, G. Short-term spatio-temporal forecasting of photovoltaic power production. IEEE Trans. Sustain. Energy 9, 538–546. https://doi.org/10.1109/TSTE.2017.2747765 (2018).
Article  ADS  Google Scholar 
Antonanzas, J. et al. Review of photovoltaic power forecasting. Sol. Energy 136, 78–111. https://doi.org/10.1016/j.solener.2016.06.069 (2016).
Article  ADS  Google Scholar 
Phan, Q.-T., Wu, Y.-K. & Phan, Q.-D. Tsmixer: an innovative model for advanced bias correction of nwp solar irradiance and one-day-ahead power forecasting. In 2025 IEEE/IAS 61st Industrial and Commercial Power Systems Technical Conference (I&CPS) 1–6 (2025). https://doi.org/10.1109/ICPS64254.2025.11030368.
Phan, Q.-T., Wu, Y.-K., Phan, Q.-D. & Tan, W.-S. A novel dual-focused temporal-spatial model for day-ahead pv power forecasting. In 2025 IEEE Industry Applications Society Annual Meeting (IAS) 1–5 (2025). https://doi.org/10.1109/IAS62731.2025.11061657.
Diagne, M., David, M., Lauret, P., Boland, J. & Schmutz, N. Review of solar irradiance forecasting methods and a proposition for small-scale insular grids. Renew. Sustain. Energy Rev. 27, 65–76. https://doi.org/10.1016/j.rser.2013.06.042 (2013).
Article  Google Scholar 
Zheng, X., Bai, F., Zhuang, Z., Chen, Z. & Jin, T. A new demand response management strategy considering renewable energy prediction and filtering technology. Renew. Energy 211, 656–668 (2023).
Article  Google Scholar 
Huang, S., Zhao, Z., Li, C. & Bai, L. Timekan: Kan-based frequency decomposition learning architecture for long-term time series forecasting. arXiv preprint arXiv:2502.06910 (2025).
Liu, Y. et al. itransformer: inverted transformers are effective for time series forecasting. arXiv preprint arXiv:2310.06625 (2023).
Zamo, M., Mestre, O., Arbogast, P. & Pannekoucke, O. A benchmark of statistical regression methods for short-term forecasting of photovoltaic electricity production, part i: Deterministic forecast of hourly production. Sol. Energy 105, 792–803. https://doi.org/10.1016/j.solener.2013.12.006 (2014).
Article  ADS  Google Scholar 
Markovics, D. & Mayer, M. Comparison of machine learning methods for photovoltaic power forecasting based on numerical weather prediction. Renew. Sustain. Energy Rev. 161, 112364. https://doi.org/10.1016/j.rser.2022.112364 (2022).
Article  Google Scholar 
Abdellatif, A. et al. Forecasting photovoltaic power generation with a stacking ensemble model. Sustainability 14, 11083. https://doi.org/10.3390/su141811083 (2022).
Article  CAS  ADS  Google Scholar 
Wang, F. et al. A satellite image data based ultra-short-term solar pv power forecasting method considering cloud information from neighboring plant. Energy 238, 121946. https://doi.org/10.1016/j.energy.2021.121946 (2022).
Article  Google Scholar 
Zhou, H. et al. Short-term photovoltaic power forecasting based on long short term memory neural network and attention mechanism. IEEE Access 7, 78063–78074. https://doi.org/10.1109/ACCESS.2019.2923006 (2019).
Article  Google Scholar 
Mellit, A., Pavan, A. M. & Lughi, V. Deep learning neural networks for short-term photovoltaic power forecasting. Renew. Energy 172, 276–288. https://doi.org/10.1016/j.renene.2021.02.166 (2021).
Article  Google Scholar 
Wang, Y., Liao, W. & Chang, Y. Gated recurrent unit network-based short-term photovoltaic forecasting. Energies 11, 2163. https://doi.org/10.3390/en11082163 (2018).
Article  ADS  Google Scholar 
Hu, Z., Gao, Y., Ji, S., Mae, M. & Imaizumi, T. Improved multistep ahead photovoltaic power prediction model based on lstm and self-attention with weather forecast data. Appl. Energy 359, 122709. https://doi.org/10.1016/j.apenergy.2024.122709 (2024).
Article  Google Scholar 
Li, Q. et al. A multi-step ahead photovoltaic power forecasting model based on timegan, soft dtw-based k-medoids clustering, and a cnn-gru hybrid neural network. Energy Rep. 8, 10346–10362. https://doi.org/10.1016/j.egyr.2022.08.180 (2022).
Article  Google Scholar 
Fu, H., Zhang, J. & Xie, S. A novel improved variational mode decomposition-temporal convolutional network-gated recurrent unit with multi-head attention mechanism for enhanced photovoltaic power forecasting. Electronics 13, 1837. https://doi.org/10.3390/electronics13101837 (2024).
Article  Google Scholar 
Agga, A., Abbou, A., Labbadi, M., Houm, Y. E. & Ou Ali, I. H. Cnn-lstm: an efficient hybrid deep learning architecture for predicting short-term photovoltaic power production. Electric Power Syst. Res. 208, 107908. https://doi.org/10.1016/j.epsr.2022.107908 (2022).
Article  Google Scholar 
Limouni, T., Yaagoubi, R., Bouziane, K., Guissi, K. & Baali, E. H. Accurate one step and multistep forecasting of very short-term pv power using lstm-tcn model. Renew. Energy 205, 1010–1024. https://doi.org/10.1016/j.renene.2023.01.118 (2023).
Article  Google Scholar 
Alharkan, H., Habib, S. & Islam, M. Solar power prediction using dual stream cnn-lstm architecture. Sensors 23, 945. https://doi.org/10.3390/s23020945 (2023).
Article  PubMed  PubMed Central  ADS  Google Scholar 
Zhou, H. et al. Informer: beyond efficient transformer for long sequence time-series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35 11106–11115 (2021). https://doi.org/10.1609/aaai.v35i12.17325.
Wu, H., Xu, J., Wang, J. & Long, M. Autoformer: decomposition transformers with auto-correlation for long-term series forecasting. In Advances in Neural Information Processing Systems, vol. 34 (eds. Ranzato, M. et al.) 22419–22430 (Curran Associates, Inc., 2021).
Zhou, T. et al. FEDformer: frequency enhanced decomposed transformer for long-term series forecasting. In Proceedings of the 39th International Conference on Machine Learning , vol. 162 of Proceedings of Machine Learning Research (eds. Chaudhuri, K. et al.) 27268–27286 (PMLR, 2022).
Moon, J. A multi-step-ahead photovoltaic power forecasting approach using one-dimensional convolutional neural networks and transformer. Electronics 13, 2007. https://doi.org/10.3390/electronics13112007 (2024).
Article  Google Scholar 
Xu, S., Zhang, R., Ma, H., Ekanayake, C. & Cui, Y. On vision transformer for ultra-short-term forecasting of photovoltaic generation using sky images. Sol. Energy 267, 112203. https://doi.org/10.1016/j.solener.2023.112203 (2024).
Article  Google Scholar 
Zhai, C. et al. Photovoltaic power forecasting based on vmd-ssa-transformer: multidimensional analysis of dataset length, weather mutation and forecast accuracy. Energy 324, 135971. https://doi.org/10.1016/j.energy.2025.135971 (2025).
Article  Google Scholar 
Deng, R. et al. A high-precision photovoltaic power forecasting model leveraging low-fidelity data through decoupled informer with multi-moment guidance. Renew. Energy 250, 123391. https://doi.org/10.1016/j.renene.2025.123391 (2025).
Article  Google Scholar 
Hu, K. et al. Short-term photovoltaic power generation prediction based on copula function and cnn-cosattentiontransformer. Sustainability 16, 5940. https://doi.org/10.3390/su16145940 (2024).
Article  ADS  Google Scholar 
Li, Y. et al. Interpretable deep learning framework for hourly solar radiation forecasting based on decomposing multi-scale variations. Appl. Energy 377, 124409 (2025).
Article  Google Scholar 
Somvanshi, S., Javed, S. A., Islam, M. M., Pandit, D. & Das, S. A survey on kolmogorov-arnold network. ACM Comput. Surv. 58, 1–35. https://doi.org/10.1145/3743128 (2025).
Article  Google Scholar 
Dutta, A. et al. The first two months of kolmogorov-arnold networks (kans): a survey of the state-of-the-art. Arch. Comput. Methods Eng. https://doi.org/10.1007/s11831-025-10328-2 (2025).
Article  Google Scholar 
Toscano, J. D., Wang, L.-L. & Karniadakis, G. E. Kkans: Kurkova-kolmogorov-arnold networks and their learning dynamics. Neural Netw. 191, 107831 (2025).
Article  PubMed  Google Scholar 
Vaca-Rubio, C. J., Blanco, L., Pereira, R. & Caus, M. Kolmogorov-arnold networks (kans) for time series analysis. In 2024 IEEE Globecom Workshops (GC Wkshps) 1–6 (IEEE, 2024). https://doi.org/10.1109/gcwkshp64532.2024.11100692.
Xu, K., Chen, L. & Wang, S. Kolmogorov-arnold networks for time series: bridging predictive power and interpretability (2024). arXiv:2406.02496.
Koenig, B. C., Kim, S. & Deng, S. Kan-odes: Kolmogorov–arnold network ordinary differential equations for learning dynamical systems and hidden physics. Comput. Methods Appl. Mech. Eng. 432, 117397. https://doi.org/10.1016/j.cma.2024.117397 (2024).
Article  MathSciNet  Google Scholar 
Download references
This research was funded by the National Natural Science Foundation of China, grant number 51967004.
This research was funded by the National Natural Science Foundation of China, grant number 51967004.
College of Electrical Engineering, Guizhou University, Guiyang, China
Linjie Liu, Min Liu, Zhuangchou Han & HaiQiang Zhao
North Alabama International College of Engineering and Technology, Guizhou University, Guiyang, China
Min Liu
Guizhou Provincial Key Laboratory of Power System Intelligent Technologies, Guiyang, China
Min Liu
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
L.L. (first author) conceived the research idea, developed the MKAN-iTransformer model, implemented all experiments, and wrote the initial draft of the manuscript. M.L. (corresponding author) supervised the entire research process, provided key guidance on model design and result analysis, and substantially revised the manuscript. Z.H. and H.Z. contributed to data preprocessing, experimental support, and manuscript review. All authors have read and approved the final version of the manuscript.
Correspondence to Min Liu.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Liu, L., Liu, M., Han, Z. et al. Interpretable ultra-short-term photovoltaic power forecasting with multi-scale temporal modeling and variable-wise attention. Sci Rep 16, 10336 (2026). https://doi.org/10.1038/s41598-026-39797-6
Download citation
Received: 01 August 2025
Accepted: 09 February 2026
Published: 24 February 2026
Version of record: 27 March 2026
DOI: https://doi.org/10.1038/s41598-026-39797-6
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Scientific Reports (Sci Rep)
ISSN 2045-2322 (online)
© 2026 Springer Nature Limited
Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.

source

Many countries and regions are gradually facing a situation where early-installed photovoltaic modules are reaching the end of their service life; this establishes a solid market foundation for the establishment of solar module recycling plant. For enterprises planning to invest in the solar module recycling sector, selecting suitable country in which to site a facility is a critical factor for project success.
Based on the scale of the solar PV market and the associated recycling demand, the following countries are particularly worthy of consideration. These countries possess substantial installed solar capacity or rapidly growing industries, providing a stable supply of raw materials and market demand for the establishment of solar module recycling plants.
Europe
Many European countries were early adopters of photovoltaic (PV) technology and possess a substantial installed base; consequently, they are poised to gradually enter a peak phase for solar module decommissioning over the coming years. The European Union features a well-developed regulatory framework for waste management—including explicit requirements for e-waste recycling—which provides a solid policy foundation for the establishment of solar module recycling plant.
Countries such as Germany, Spain, and Italy—which rank among the leaders in PV installed capacity and possess mature supporting industrial ecosystems—are particularly well-suited locations for siting solar module recycling projects.
China
China stands as the world’s largest producer and installer of solar PV systems, boasting a cumulative installed capacity exceeding 500 GW. This signifies that a massive volume of end-of-life solar panels will be generated in the coming years, making the establishment of recycling facilities an effective means of processing discarded solar modules locally. Concurrently, relevant government authorities have recently begun actively promoting the development of standards and pilot programs specifically dedicated to solar panel recycling.
Selecting China as the location for solar module recycling plant allows for the full leveraging of local supply chain advantages, thereby ensuring the highly efficient operation of the recycling process.
United States
The U.S. solar market continues to grow, bolstered by strong support for renewable energy across various states. Some states have already begun to address the issue of end-of-life PV modules and are advancing relevant legislation. For instance, federal and state-level policies—such as the “Infrastructure Investment and Jobs Act”—provide financial support for recycling initiatives.
Establishing solar module recycling plant within the United States helps ensure compliance with domestic environmental regulations, offers manufacturers localized recycling services, and reduces transportation costs.
India
India’s installed solar capacity is climbing rapidly, having already surpassed 50 GW, with a government target of reaching 450 GW by 2030. This trajectory will generate a massive demand for recycling services.
Although India is an emerging market for photovoltaics—and its recycling infrastructure is currently still under development—proactively establishing solar panel recycling facilities will enable the country to meet future market demands while simultaneously fostering job creation and local technological advancement.
Australia
 
Australia boasts a high rate of solar energy adoption, alongside a steady increase in utility-scale ground-mounted solar projects. The government is actively promoting renewable energy recycling initiatives—such as the Solar Recycling Fund—which offer incentives for the construction of new recycling plants.
Setting up solar panel recycling plants in the vicinity of Australia‘s major cities or industrial zones would enable coverage of both local markets and select markets within the Pacific region.
These countries hold significant potential for solar panels recycling, as their market scale and policy environments are conducive to supporting the long-term operation of solar panel recycling plant. Naturally, specific investment decisions should be made in conjunction with local regulations and economic assessments. If you are considering investing in solar panel recycling plant in the aforementioned countries or other regions, we invite you to consult with the professional manufacturer of solar PV panel recycling machine: Henan DOING Company.
DOING Company’s solar panel recycling plants offer complete recycling production lines, encompassing dismantling, crushing, and sorting equipment. Utilizing the automated crushing and sorting system, our technology separates high-purity valuable materials—such as glass, silicon, aluminum, and copper—from discarded solar modules. With an extraction rate exceeding 98%, our solutions help recycling facilities minimize operational costs while ensuring compliance with international environmental standards.
Whether you plan to establish solar module recycling plant in Germany, Spain, Italy, China, the United States, India, or any other country, our equipment can be tailored to meet local requirements, and we are ready to provide customized recycling solutions to suit your specific needs.
Environmental XPRT is a global environmental industry marketplace and information resource. Online product catalogs, news, articles, events, publications & more.
Channels
Solutions
Latest
Publications
Companies

source

Our name is about honesty. The Monitor is owned by The First Church of Christ, Scientist, and we’ve always been transparent about that.
The church publishes the Monitor because it sees good journalism as vital to progress in the world. Since 1908, we’ve aimed “to injure no man, but to bless all mankind,” as our founder, Mary Baker Eddy, put it.
Here, you’ll find award-winning journalism not driven by commercial influences – a news organization that takes seriously its mission to uplift the world by seeking solutions and finding reasons for credible hope.
Get unlimited access to global news that cuts through the noise and brings perspectives you won’t get anywhere else.
Already a subscriber?  Log in
Already have a subscription?  Activate it
Join the Monitor community.
Already a subscriber? Log in to hide ads.
Loading…
April 04, 2026, 5:00 a.m. ET
The solar industry became a target in the first year of President Donald Trump’s second term, when his administration cut key federal tax credits, subsidies, and investments in solar power, as well as broader green technology initiatives.
The Republican president’s skepticism toward low-carbon energy, rooted in a combination of economic, aesthetic, and ideological objections, is well-known. At the World Economic Forum held in January 2025, Mr. Trump boasted of terminating former Democratic President Joe Biden’s “ridiculous” and “wasteful” Inflation Reduction Act – which, among other spending to support renewable energy, offered tax incentives to encourage solar power.
Yet as America’s demand for electricity rises – expected to grow at a 2.8% compound annual growth rate over the next 15 years – some Republican influencers such as former House Speaker Newt Gingrich and former Trump adviser Kellyanne Conway, who are still among the president’s most ardent supporters, are encouraging him to adopt a more pragmatic approach to solar.
America’s rising demand for electricity is putting focus on where that energy should come from. As President Donald Trump has sought to elevate fossil fuels, one fast-growing renewable energy source is attracting support from an unlikely place: political conservatives.
The reason is simple supply and demand, they say, though Mr. Gingrich mixes in a touch of nationalism. Failure to meet the electricity demands of industry would slow business growth, just when futuristic technologies such as artificial intelligence are starting to boom. To remain competitive with China in AI, he says, America needs to expand its electric power generation by using every energy source available.
“History tells us that energy scarcity is the biggest threat to the American economy,” Mr. Gingrich wrote in a recent opinion column in the right-leaning website, the Daily Caller.
“We need more of everything. Intentionally excluding vital energy sources, fuel-based or renewable, reduces supply and drives up prices. This harms families and businesses. This is not abstract economic theory. It is common sense.”
Historically, America’s power grid has always been a pragmatic concern, though people at opposite ends of the political spectrum often disagree on how to ensure Americans get the power they need.
The debate over climate change disrupted all that.
In 2016, America, under President Barack Obama, joined 192 other nations and the European Union in signing the Paris Agreement to substantially reduce carbon emissions by 2030. Americans were divided: Conservatives said the agreement would damage America’s economy, while liberals cited the overwhelming scientific consensus that climate change is happening and is primarily caused by human activity. How and whether to use low- and zero-carbon energy sources such as wind and solar – moving away from fossil fuels – were at the heart of that partisan divide.
But today, even amid disagreements, after a decade and a half of expansion, solar power accounts for 8.5% of the U.S. electricity generation mix, up from 0.1% in 2010. That is still well behind fossil fuels, which make up more than half of the U.S. energy mix. Fossil fuels have a powerful lobbying presence, spending about $150 million a year to advocate for oil, gas, and coal businesses. Even so, in 2024, wind and solar power together overtook coal for the first time in electricity production, with wind and solar at 17% and coal at 15%.
“I don’t view energy as an ideological issue, but that’s what it’s been turned into,” says Eric D. Larson, a senior research scholar at the Andlinger Center for Energy and the Environment at Princeton University.
“I think that maybe a concern is that if demand is outstripping supply, prices will go up,” Professor Larson says. “And rising prices translate into votes, typically. But also, there is the recognition that if we want to stay ahead as a country at the cutting edge of technology, and AI is going to be a big part of that, we need power.”
At the moment, there is little evidence that prominent Republicans touting solar will influence President Trump to change his mind on solar energy. His 2025 tax-and-spending bill, which phased out solar subsidies, passed the House along party lines, 218-214. Those entrusted with implementing Mr. Trump’s energy policies are led by Secretary of Energy Christopher Wright, the founder and former CEO of Liberty Energy, a major fracking service company. He has called solar panels “a parasite” that provides only intermittent power. In a Feb. 17, 2026, panel discussion, Mr. Wright showed no signs of wavering.
“We got off track because of a wild misunderstanding, an exaggeration” about climate change, Mr. Wright said at a conference at the French Institute of International Relations in Paris. Climate change, he said, “is a real thing, but it has gotten so ridiculously out of whack that we have policies that have just driven up energy [prices, and] driven deindustrialization and made our countries geopolitically weaker.”
Even so, a small number of Republicans with influence in the White House are publicly advocating that President Trump adopt it as part of “America First” policy and give solar energy another chance.
Conservative podcaster Katie Miller – wife of Mr. Trump’s Homeland Security Advisor Stephen Miller – recently made the case for solar energy on social media. Ms. Miller, a former aide to Elon Musk, has been publicly promoting solar energy and has noted that her former boss’s company, Tesla, produces solar panels. She has stated she does not have a paid partnership with the clean energy groups she sometimes cites.
“Solar is now the dominant source of new U.S. power capacity and is on track to surpass coal in total installed capacity before the end of 2026,” she wrote. “70 GW of new solar capacity is scheduled to come online in 2026–2027 → a 49% increase in operating solar capacity from the end of 2025.”
Ms. Conway, a pollster and former Trump senior political counselor, conducted a February 2026 poll of 1,000 registered voters in Arizona, Florida, Indiana, Ohio, and Texas, on behalf of the pro-solar advocacy group, American Energy First. Her findings: “Solar power enjoys broad, durable, and increasingly intense public support,” including among Trump voters.
Eight in 10 respondents agreed when asked if “solar energy should be used in the U.S. to strengthen and increase our energy supply?” Three-quarters of self-identified Trump voters agreed.
More than two-thirds of the new solar plants built over the past five years are in states that the Republican Party carried in 2024, including Texas, Indiana, Florida, Ohio, Arizona, Utah, and Arkansas, according to a report by Wood Mackenzie on behalf of the Solar Energy Industries Association.
Artificial intelligence is booming, “and if you look at where they are building data centers, it’s in red-state America,” says Mark Fleming, president and CEO of Conservatives for Clean Energy in Raleigh, North Carolina. In the Carolinas, the combination of AI data centers and solar and wind farms has contributed to the property tax base. “It’s been a lifesaver for rural counties,” Mr. Fleming says, and that has brought more conservatives to the clean energy cause.
Neil Auerbach, founder and CEO of the Hudson Sustainable Group and senior adviser to the conservative-leaning American Conservation Coalition, says that ideologues on both sides of the aisle miss the point. America’s economic competitiveness depends on “abundant, affordable power today.”
Deepen your worldview
with Monitor Highlights.
Already a subscriber? Log in to hide ads.
“With the 2026 midterms approaching, the political stakes are clear. Voters will reward leaders who present credible solutions to keep the lights on and bills down,” Mr. Auerbach wrote in a recent opinion column. “Ideological purity, whether hostility toward fossil fuels or toward renewables, will not deliver affordable energy. An all of the above approach is not ideology; it is economic realism.”
Samantha Gross, director of the Energy Security and Climate Initiative at the Brookings Institution, says that whether you are worried about climate change or not, solar energy is a “one of the easiest and fastest ways to get power. That is one reason why the administration shouldn’t be rolling its eyes at solar power.”
Already a subscriber?  Log in
Monitor journalism changes lives because we open that too-small box that most people think they live in. We believe news can and should expand a sense of identity and possibility beyond narrow conventional expectations.
Our work isn’t possible without your support.
Already a subscriber? Log in
ALREADY A SUBSCRIBER?  Login

Link copied.
We want to hear, did we miss an angle we should have covered? Should we come back to this topic? Or just give us a rating for this story. We want to hear from you.
Already a subscriber? Log in to hide ads.
Follow us:
Your subscription to The Christian Science Monitor has expired. You can renew your subscription or continue to use the site without a subscription.
Return to the free version of the site
If you have questions about your account, please contact customer service or call us at 1-617-450-2300.
This message will appear once per week unless you renew or log out.
Your session to The Christian Science Monitor has expired. We logged you out.
Return to the free version of the site
If you have questions about your account, please contact customer service or call us at 1-617-450-2300.
You don’t have a Christian Science Monitor subscription yet.
Return to the free version of the site
If you have questions about your account, please contact customer service or call us at 1-617-450-2300.

source

By the People, for the People
News
The 5-megawatt community solar project will bring renewable energy to the local area.
Mar. 31, 2026 at 11:55am
Got story updates? Submit your updates here. ›
Doña Ana County commissioners have approved up to $10 million in industrial revenue bonds to support the development of a new 5-megawatt community solar farm in the town of Vado, New Mexico. The solar project is expected to bring renewable energy access and economic benefits to the local community.
This investment in community-scale solar power aligns with broader efforts in New Mexico to expand renewable energy generation and access, particularly in underserved areas. The Vado solar farm will provide clean electricity to local residents and businesses, helping to reduce reliance on fossil fuels and lower energy costs.
The $10 million in industrial revenue bonds will help finance the construction and operation of the 5-megawatt solar farm in Vado. The project is being developed by a local renewable energy company and is designed to supply clean electricity to homes and businesses in the surrounding community through a community solar program.
The local county government that approved the bond financing for the Vado solar project.
The town in Doña Ana County where the 5-megawatt community solar farm will be located.
With the bond financing approved, the solar project developers will now move forward with securing permits, finalizing contracts, and beginning construction on the Vado community solar farm.
This investment in local renewable energy infrastructure demonstrates Doña Ana County’s commitment to expanding access to clean, affordable electricity for its residents and businesses through innovative community solar programs.
We keep track of fun holidays and special moments on the cultural calendar — giving you exciting activities, deals, local events, brand promotions, and other exciting ways to celebrate.

source

AIKO has announced a new collaboration with Solar Team Eindhoven, bringing its high-efficiency ABC (All Back Contact) solar cells to power Stella Juva, the world’s first solar-powered ambulance designed to operate entirely on solar energy while supporting onboard medical equipment.
Developed by students from the Eindhoven University of Technology, Stella Juva aims to enable healthcare delivery in remote or infrastructure-limited regions. Expected to hit the road in July 2026, the project represents a significant step in redefining the role of solar-powered mobility, from transport solutions to mobile energy systems supporting essential services.
David Komdeur, Solar Team Photovoltaics Engineer, commented: “Stella Juva pushes the boundaries of what solar technology can achieve in real world applications. We chose AIKO as our partner because of its industry leading efficiency and proven reliability, both critical for a vehicle that must operate independently under varying conditions. The ABC cells use a full back contact design without front side metallization, which maximizes light absorption. In addition, the silver free metallization lowers the risk of microcracks and contributes to long term durability. A low temperature coefficient, combined with strong resistance to degradation, helps maintain stable performance across a wide range of environments.”
The partnership marks a further expansion of AIKO’s engagement with leading solar mobility teams, moving beyond competition platforms toward real-world applications of zero-carbon mobility and sustainable living. By integrating ABC technology into a functional emergency vehicle, the collaboration demonstrates how advanced photovoltaic innovation can contribute to both clean transportation and critical societal needs. For AIKO, this reflects a broader commitment to enabling high-efficiency solar technology in emerging application scenarios, supporting projects that integrate energy generation directly into mobility and infrastructure and exploring how photovoltaic innovation can deliver value beyond traditional installations.

source

MnSEIA, SEIA, And CCSA Criticize Minnesota’s Approval Of Xcel’s 200 MW Battery Program, Warning Of Risks To Ratepayers  SolarQuarter
source

Global photovoltaic waste under ratcheting climate ambition: Spatio-temporal distribution and future pathways  ScienceDirect.com
source

Local
News
News
News
News
News

source