It’s common for Passive House homes to have solar panels on their roofs. Every Passive House Plus or Passive House Premium project has to generate at least as much energy as its household plug load over the course of a year and in New Zealand this has been done via photo-voltaic (PV) panels. I want people to be clear that off-grid PV systems are not inherently a good choice or the pinnacle of PV generation that sustainable minded folks should aim for. If this seems like a provocative idea, please keep reading. I’ll outline the science that backs up this statement.
I’m evaluating this through a carbon lens, which is something of a theme for the month as we get ready to release The Carbonator, the tool we built to plug into PHPP and help Passive House designers evaluate the carbon costs associated with different materials and levels of specification. Obviously PV panels make positive contributions to the operational emissions associated with a house (or any other building). But generating your own electricity on site comes at a significant upfront cost—that is, in embodied carbon.
There are a lot of CO2 emissions associated with manufacturing PV panels. Some of this is from the metals used but very largely it’s the silicon wafers. The silicon must be extracted, purified, grown and sliced; all these activities are energy intensive. This was graphically illustrated for our team while we worked on the Fletcher Living LowCO project, an ambitious attempt to produce mainstream home designs with 80% smaller carbon footprints.
PV panels did pass the rigorous screen Fletcher LIving applied and they were specified for both the stand alone home and the terrace houses. The technology and manufacturing has improved in leaps and bounds and it is only modern panels that can produce a net carbon benefit. Twenty years ago, there wasn’t a hope of the operational carbon benefit outweighing the embodied carbon cost.
Whole of building carbon impact over 50 year service life. © 2024 Sustainable Engineering Ltd
For the LowCO detached house, we estimated 25,250 kgCO2-eq embodied carbon was emitted while 31,241 kgCO2-eq operational carbon savings was won from energy that it generated that was used in the home and exported to the grid over 50 years. This is fairly balanced; it could be considered a 25% return on embodied carbon invested. Industry expects the embodied carbon in PV systems to continue to reduce and this is critically important*.
This won’t be an important consideration for some people. But our company is seeing increasing numbers of clients whose primary motivation is to build responsibly for the planet and who are asking us hard questions about the embodied carbon emissions associated with higher levels of performance. That’s what led us to develop our own means of quickly comparing different materials and performance levels. We started with insulation quantities, seeking the sweet spot where operational carbon savings are optimised and embodied carbon costs are minimised. We then moved on to investigate other components. (Read more about the tool we playfully dubbed The Carbonator.)
Taking context into account makes everything both more complicated and more accurate. Given New Zealand’s relatively clean grid, the question a design team should ask is the operational carbon savings due to the electricity produced by the PV enough to more than compensate for the large amount of embodied carbon in the panels and mounting structure?
The analysis done to estimate the carbon impacts assumes that every watt of energy generated by the panels (on that specific site) is used. There are inherent inefficiencies in an off-grid PV set up. In order to have enough electricity in winter, there is invariably far more power generated in summer than can be used. It leads to perverse outcomes like dumping energy into resistance, hot water heating etc—essentially finding a way to waste it.
Modern systems are a bit more sophisticated and inverters shut down the panels’ capacity to generate energy but it amounts to the same thing: the actual amount of energy generated for a useful purpose is less than the theoretical maximum amount of energy capable of being generated. This means less operational carbon benefit in reality, which makes it harder to reach a net carbon benefit. We can see from the actual analysis of the LowCO example that the carbon cost and carbon benefit numbers aren’t so far apart; if all the energy capable of being generated is not generated and used, the system could end up as carbon positive. That’s a bad thing, remember: more carbon emitted than carbon saved.
I acknowledge that there are instances for remote sites where it is not possible to connect to the grid, because of physical distance from infrastructure or prohibitive costs or both. Costs may be high in dollar terms and also with regard to the carbon emissions associated with the physical infrastructure. In these cases it is very important to absolutely minimise the power needed and therefore the size of the PV system and battery required.
In all other cases, I strongly recommend that homeowners invest in grid-tied systems and use the grid as their battery. This ensures all the energy the system is capable of generating is collected and distributed for use (your neighbour’s carbon footprint will reduce, especially when the grid is being topped up by coal-fired plants). For those concerned about keeping the fridge running and computers and internet on during a protracted black out, investigate the options that now exist for plugging an electric vehicle into your home and using its battery charge in the home. That’s two purposes for a battery for the price of one. This article explains some of the various ways an EV battery can interact with wider systems.
*Reference: Ito, Masakazu. (2011). Life Cycle Assessment of PV systems. 10.5772/23134.