Are Solar Batteries Zero Emissions?
At the start of August my new home battery was installed and connected to the rooftop panels I’ve had for several years. As for hundreds of thousands of Australian home-owners before me (too bad for the proportion of Australians who rent) the reasons were predominantly financial. I have little hope of seeing lower retail power prices in South Australia - or better energy policy at the national level - any time soon, and so the roughly six year payback period looked to be worth the cash investment.
But there’s another kind of payback which goes practically ignored: how much does clean, green solar power, stored into a battery and then discharged later after dusk, truly help the environment and the climate? Many would ignore it because they assume the answer is “loads and loads” - so much that the question might even seem stupid. And that’s fine when you’ve never thought about the amounts of materials used for making things like PV panels and batteries, and the amount of mining needed for producing those materials. Most of these people have other things to worry about in their lives.
In contrast, certain other stakeholders have lots to say about how environmentally and culturally harmful mining can be. But it isn’t guided by frank analysis of the best available data, as found in the comprehensive 2024 work out of The Breakthrough Institute, above. Instead, the mining and material intensity of specific “undesirable” energy technologies have been emphasised - and sometimes exaggerated - just to refuse them the badge of zero emissions. This isn’t a new argument, and for nuclear energy in particular, it stubbornly endures.
As an Australian who likes solar panels and battery storage so much that I’ve installed both, I reckon it’s time for them to be tested with the same logic chosen by their lobby groups.
The important metric to understand is the grams of carbon dioxide-equivalent per kilowatt hour (gCO₂eq/kWh ) estimated for each electrical energy source. Individual values are very low for nuclear and renewable technologies, high for gas and coal. No power source is 0 gCO₂eq/kWh - even if operating it produces no smoke - due to the emissions-intensive industrial processes required in its full lifecycle. Regional and national power grids, composed of mixed energy sources, also exhibit gCO₂eq/kWh values which change over time - visualised best at Electricity Maps.
So, solar power isn’t “zero emissions” but it’s still pretty low over its full lifecycle. The climate benefit from displacing fossil fuels is clear. Storing it in batteries just means it can displace them at night, too… is the accepted wisdom, without genuinely considering the combined solar+battery gCO₂eq/kWh.
What might this number be? Work on this has been rather limited, but a 2022 review article in the Journal of Energy Storage provides a good place to start.
My new battery system is based on lithium iron phosphate (LFP) technology manufactured by Sigenergy, so the reviewed literature suggests a range of around 90 to nearly 200 gCO₂eq for a typical kilowatt hour I’ll use from my system. Certainly not “zero emissions”: to paraphrase The Climate Council, “greenhouse gas pollution associated with solar and batteries could be similar to a gas power station with carbon capture & storage”.
I’d rather be at (or better, below) the lower end, but the analysis concludes that (a) embodied emissions for LFP are usually higher than other options, and (b) the exact value mostly depends on the emissions intensity of the power to the manufacturing facility, offset by the proportion of natural gas used for heating processes. In other words, it looks like higher proportions of gas-fired process heat save on emissions when LFP batteries are produced in locations with “dirtier” electric grids.
Sigenergy produces its batteries in Shanghai, and recent analysis notes that such facilities rely on the local coal-dominated, emissions-intensive grid for 98% of their energy consumption. A lifecycle analysis from last year estimated 93.1 kg of CO₂eq per kWh capacity for LFP battery production in factories like Sigenergy’s. This is marginally lower than the most recent datapoint in Figure 4 above, but significantly higher than the low values ("Wang", "GREET") for LFP.
Referring back to Figure 3 with this estimate in mind, my solar+battery system may unfortunately be up above 100 gCO₂eq/kWh over its operational life. I don’t think this is overestimated; in contrast, the highest value for LFP (based on Oliveira et al. 2015) traces back to a derived manufacturing CO₂eq per kWh footprint just one fifth of the Shanghai estimate, thus effectively underestimating the total lifecycle emissions. How high would the bar reach on that figure? Past the United Kingdom? Past natural gas? If we could determine the correct assumptions for my PV panels and battery, just how much higher might it loom over that dotted line for Nuclear down there closest to the axis?
If the overall emissions intensity of the power from my battery is higher than South Australia’s average gCO₂eq/kWh, then maybe trying to save a bit on power bills over time is the only real reason to get one anyway.
Organisations like The Climate Council regularly promote solar and batteries as infinitely more viable climate solutions than nuclear energy. They also report millions of dollars of charity revenue. A little of that could fund the specific research needed to determine how all the steps involved in producing battery power in Australia at night - from mining, to construction, decommissioning and waste management - result in greenhouse gas pollution. Seems pretty important, and maybe The Climate Council doesn’t really want an answer. But I do. Do you?