We’ve been reviewing previous work on the life cycle emissions for photovoltaics. Below are a few of the more relevant studies we’ve found. However we have not really come across a recent authoritative literature review / meta-analysis. Are you aware of any such summaries, especially regarding the current market reality? We’re also seeing that Wp and kWh are both used as functional units, depending on the PCR / program. How does using kWh make any sense!? Any help much appreciated!
Some background on this request: To evaluate PV impact, we have been projecting 25-years with Cambium LRME at year 1, followed by AER in subsequent years. We’re using the hourly emissions factors to better estimate avoided GHG (TDV). The EPD’s we’ve found for PV range from 0.3 to 0.8 kgCO2e/Wp. Putting it together, in places like Colorado we’re seeing >20yr GHG payback. However we expected to see between 1 and 7 years… Hence the quest for better understanding.
It’s an interesting question, but I wonder if reframing it might be useful?
PV is the cheapest source of energy in many parts of the US and we need to build a great deal of new energy generation to supply the EV market, industry, and electrification of buildings (plus batteries/storage and transmission lines and more). So we are going to build a great deal more solar energy using PVs, plus wind, geothermal, etc. and many states are requiring utilities to do just that. Interestingly enough, even states that haven’t required utilities to build wind and solar are doing so because they are cheap energy sources. The US building a lot more solar is a baseline case within Cambium to reflect this, and my understanding is that Cambium does not yet fully capture the Inflation Reduction Act acceleration of the renewable energy transition.
So if lots of PV getting built is a baseline case, does that change your question? Is it about whether each PV panel has a payback against a decarbonizing grid (that is made of up of other new PV panels and their embodied carbon), is it about whether you build the PV or someone else does, or is it about how PV panels’ embodied carbon and operational carbon compare to embodied and operational carbon from wind, nuclear, or gas-fired power plants?
@stuart.s - great question! In case it’s a useful addition to your list, I’d encourage you to check out this report that Louise Hamot and some of my other colleagues put out last year in partnership with Wilmott Dixon.
Whole Life Carbon of PV
Figure 5 in particular may be of help to you if you’re focused on just the panels.
PS: I agree that Wp seems to be the better functional unit (as the kwh per panel may change depending on solar intensity per region).
PPS: and yes, depending on the grid we’re seeing large ranges in terms of GHG payback times. For projects in hydropowered BC, I’m not sure it ever pays back (the PV reaches EOL before paying off the carbon premium…)
and @Kjell_Anderson - agreed - there’s many ways to frame that question. I personally have been trying to think about how PV panels’ embodied carbon and operational carbon compare to embodied and operational carbon from wind, nuclear, or gas-fired power plants (alongside their generally much more comprehensive distribution infrastructure)
I’m optimistic I’ll be able to publish another case study in the not too distant future on the topic - when that happens, I’ll post on this thread again!
This is the only study that I’ve seen on that analysis @JeremyField. I don’t fully understand it as it is a projection for 2050. I’ve looked up the data sources and only understand the conclusions somewhat. It does include fuel and embodied carbon of the facilities. Understanding future emissions from low-carbon power systems by integration of life-cycle assessment and integrated energy modelling | Nature Energy
Thanks for sharing that @Kjell_Anderson - I’ll check it out!
Indeed the power you are supplanting makes a big difference, for me in BC it would be better to fund a solar panel installation in India than to buy my own here. One thing to bear in mind is the PV grade silicon is currently produced by reducing silica using carbon (similar to steel production), so there are process emissions that can only be eliminated by moving to another reductant (hydrogen), or a different process route (perhaps electrolysis). Assuming we stick with silicon PV this paper examines the carbon budget required from process emissions to get to 100% renewable: https://www.sciencedirect.com/science/article/abs/pii/S2542435123000016 (along with other resource requirements).
There is a lot of promise of perovskite solar cells that should be lower emissions in production, but I believe the durability is not yet sufficient.
Thank you @Kjell_Anderson @JeremyField and @will.nash - I’m humbled by the expertise and great literature references - my understanding of the ‘trade-off’ of PV embodied carbon is already turned on its head.
The white paper you shared Jeremy by Louise Hamot is done wonderfully. We’re left wondering about the embodied carbon of other generation resources as the best tactic to evaluate on-site solar. It seems wind is offers the lowest up-front emissions, when amortized over the life of the system on a per kWh basis…
Although that is a different post! Questions beget questions