{"id":19759,"date":"2026-09-11T08:00:30","date_gmt":"2026-09-11T12:00:30","guid":{"rendered":"https:\/\/www.autonocion.com\/us\/?p=19759"},"modified":"2026-09-11T05:26:35","modified_gmt":"2026-09-11T09:26:35","slug":"sydney-roof-14000-plants-319-solar-panels","status":"publish","type":"post","link":"https:\/\/www.autonocion.com\/us\/sydney-roof-14000-plants-319-solar-panels\/","title":{"rendered":"A Sydney factory roof grew 14,000 native plants under its 319 solar panels and they made 11 percent more electricity over the summer than a bare array, because the leaves cool the air under the glass, and on the hottest days the roof ran 50 degrees cooler than bare aluminum"},"content":{"rendered":"

I’ve spent a good chunk of this summer writing about people putting solar panels in odd places. Upright in the Arctic. Floating on an Alpine reservoir<\/a> that gets drained every year. Ten feet over a Massachusetts sweet corn field<\/a>. The pitch is usually the same, which is that you’re getting power out of ground that’s already busy doing something else.<\/p>\n

A roof in Western Sydney ran that the other way around. The panels went up first, and more than 14,000 native plants went in underneath them. Researchers from the University of Technology Sydney<\/a> then spent four months of the 2025 Australian summer measuring whether the plants did anything for the electricity. They did, by about 11 percent.<\/p>\n

So why would a plant make a solar panel work harder?<\/p>\n

Because solar panels don’t like heat. They’re semiconductors, and a hot semiconductor loses voltage. The sunlight is what you want and the temperature that comes with it is what you don’t, which is why your rooftop array is quietly slacking off on the exact August afternoon your air conditioning is working its hardest. UTS puts the sweet spot at 77 degrees Fahrenheit (25\u00b0C) and says efficiency drops above it.<\/p>\n

Plants handle heat by sweating, more or less. Water moves up through the plant and evaporates off the leaves, and evaporation pulls heat out of the air sitting right under the panels. Do that across 14,000 plants and you’ve cooled the little pocket of air the whole array is standing in.<\/p>\n

The 23.25 percent number needs an asterisk<\/h2>\n

UTS came out with two figures and they’re already being used interchangeably, which they shouldn’t be. Across the full four months, the biosolar panels beat a conventional array by an average of 11.1 percent. On the hottest, brightest days that gap opened up to 23.25 percent. The second one is a peak, not a season, and the researchers told pv magazine<\/a> the gains tracked how much light was hitting the roof that day.<\/p>\n

They’ve also put a per-panel figure on it: 0.24 kWh a day of extra generation, averaged over the study. There are 319 panels up there, so I ran that out and got roughly 77 kWh a day across the array, or somewhere around 9,200 kWh over four months. That’s about ten months of electricity for an average American home, which the EIA<\/a> puts at 10,500 kWh a year. UTS didn’t publish a season total, so that’s my arithmetic, not theirs.<\/p>\n

The heat number got me more than the electricity did. On extreme heat days the planted roof measured up to 50 degrees Fahrenheit cooler (28\u00b0C) than the conventional aluminum roof it was compared against. That’s a surface reading. No one is claiming the suburb dropped 50 degrees.<\/p>\n

One correction while we’re here, because the pictures make it look denser than it is. The array covers about 30 percent of the 14,000-square-foot roof (1,300 m\u00b2). This is mostly garden with solar panels sitting in it, not a slab of glass with a few weeds underneath.<\/p>\n

\n
\n
MEASURED<\/div>\n
SUMMER AVERAGE<\/div>\n
+11.1%<\/div>\n
Generation over a conventional array, across four months.<\/div>\n<\/div>\n
\n
HOTTEST DAYS<\/div>\n
+23.25%<\/div>\n
Peak gain during extreme heat, not a seasonal figure.<\/div>\n<\/div>\n
\n
SURFACE TEMPERATURE<\/div>\n
50\u00b0F cooler<\/div>\n
Planted roof against bare aluminum on extreme heat days (28\u00b0C).<\/div>\n<\/div>\n
\n
PER PANEL, PER DAY<\/div>\n
0.24 kWh<\/div>\n
Extra output per panel, averaged over the study. 319 panels installed.<\/div>\n<\/div>\n<\/div>\n

The roof isn’t only doing solar<\/h2>\n

The roof handles rain too. Across major rainfall events it held onto an average of 73 percent of the water that landed on it, and in five of six of those events it held all of it. The growing medium and the drainage layer soak it up instead of dumping it into a storm drain two minutes after the storm starts.<\/p>\n

It also filters. UTS measured 99.9 percent of the manganese, 99.8 percent of the nickel, 98.7 percent of the arsenic and 91.75 percent of the zinc stripped out of the runoff, and lead researcher Peter Irga described the roof as a giant natural filter. The plants also pull roughly 2.3 ounces of air pollutants out of the air each day (65 grams), which comes to about 53 pounds a year (24 kilograms).<\/p>\n

Then there’s what moved in. Field surveys and eDNA sampling turned up 39 species of birds, insects, mammals and gastropods, 13 of them pollinators, including a blue-banded bee that Irga singled out as rare. Kookaburras showed up too. Irga called the speed of it remarkable, and on a roof that’s only been planted a couple of years, I’d agree with him.<\/p>\n

Could you do this on your own roof?<\/h2>\n

Probably not as a retrofit, and that’s the honest answer. Hassell designed this roof garden with a green roof outfit called Three Owls Landscaping, and the panel layout was set up to work with the planting rather than get bolted on over the top of it. The team behind an earlier Sydney biosolar study at Barangaroo<\/a> made the same point about their own site.<\/p>\n

Weight and water are the other two problems. You’re asking a structure to carry growing medium, plants and an irrigation and drainage system, all of it soaking wet, in the middle of a storm. UTS hasn’t published what the Bradfield roof cost to build, and I’m not going to guess at it.<\/p>\n

The Barangaroo study is the obvious comparison. Two near-identical buildings, one plain solar roof, one with panels over a green roof, and the planted one came out 3.6 percent ahead. That’s a long way off 11.1 percent, and you can’t stack the two numbers side by side: Barangaroo ran eight months including winter, Bradfield ran a summer. Different question, different answer.<\/p>\n

There’s one more thing about the Bradfield site, and it cuts both ways. Irga said the data came out clean because the rest of the city hasn’t been built yet, so nothing nearby was throwing traffic exhaust or waste heat at it. Great for measurement. It also means the plants were cooling an array in a spot that doesn’t have much of an urban heat island yet. That’s the problem these roofs get sold to fix.<\/p>\n

The building itself sits in the Western Sydney Aerotropolis next to the new airport at Badgerys Creek, and it holds the first stage of an advanced manufacturing facility. Bradfield’s master plan already requires green or biosolar roofs on 80 percent of suitable roof space on new buildings there, which is a fairly aggressive thing to put in writing before anyone has measured a roof. Now someone has.<\/p>\n

UTS published the assessment on August 10, and Irga’s team is already running the same idea on prefabricated modular cabins for a developer called Future Property Group. Smaller roofs, same trick.<\/p>\n","protected":false},"excerpt":{"rendered":"

I’ve spent a good chunk of this summer writing about people putting solar panels in odd places. Upright in the … <\/p>\n

Read more<\/a><\/p>\n","protected":false},"author":8,"featured_media":19763,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[116],"tags":[],"class_list":["post-19759","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-energy","resize-featured-image"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/posts\/19759","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/comments?post=19759"}],"version-history":[{"count":3,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/posts\/19759\/revisions"}],"predecessor-version":[{"id":19764,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/posts\/19759\/revisions\/19764"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/media\/19763"}],"wp:attachment":[{"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/media?parent=19759"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/categories?post=19759"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.autonocion.com\/us\/wp-json\/wp\/v2\/tags?post=19759"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}