I used Gridcognition software to see if my house could go 100% off-grid

There’s much discussion in energy circles around how much storage will be required to deliver the zero-carbon energy systems the world is heading towards. We know we can build out renewable generation cheaply, but we also know we need storage to tide us through the periods where those (inherently variable) renewables come up short.

The same question applies at the distributed energy level. In regions where the solar resource is good, & rooftop solar is prevalent, there’s often a primeval urge to go fully off-grid, so generate 100% of your own clean energy requirements from solar & then use a combination of load flex (better aligning when you consume energy to when you produce it) & storage (battery and/or EV + V2G).

As the owner of a home with a decent sized solar system (6.5kW), fairly modest energy consumption (around 3.5MWh per year) & no storage, & inspired by related posts by Max van Someren and Mattia Marinelli, I used the Gridcognition software to have a crack at working out how easy it might be to go fully off-grid, and so 100% carbon free.

So to start with some stats and facts, my house uses about 3.5MWh of energy a year with 1.85MWh of that coming from the grid and 1.65MWh from my solar. Around 41% of the time (or 47% by volume) I operate without grid energy. In total, my 6.5kW of solar generates a bit over 10MWh a year, so I’m exporting a LOT back to the grid under ‘business as usual’. A promising prospect for storage you might think.

Now for the battery analysis, I’ve then taken the 15-minute interval data from my solar and house load and fed it into Gridcognition where I’ve modelled a range of battery sizes until I reached a point where my requirement for grid supplied energy hits zero.

So what did I find out? Here are the highlights:

• I can move from 41% carbon free hours a year to an impressive 90% with the addition of just a 3kWh battery. By volume that’s 47% to 77%.
• 95% is pretty doable too with a modest 5kWh battery required. Remember that typical residential battery sizes are in the 10kWh to 15kWh range.
• After that, things get a lot harder with 99% carbon-free requiring a 22kWh battery & then to move from 99% to 100% requires an additional 37kWh of storage. I can finally go fully off-grid with a ~59kWh battery.
• The trouble comes with is a few warm days in summer when I run the aircon a fair bit, particularly if I need to run it overnight and so it’s not aligned to solar production.

In this animation, what you’ll see is:

• Monthly Energy Balance – shows the energy imports (blue) and exports (teal) for each month of the year. I’ve excluded the gross solar generation to reduce clutter.
• Interval Data – shows grid draw (blue), solar generation (yellow) and battery state-of-charge (purple) for the month of December. I’ve used December as it includes a couple of warm days that demonstrate the challenges of right-sizing storage.
• Column Chart – just shows the percentage of grid-supplied energy for each battery size.

This is just a single year of model so battery degradation isn’t a significant factor which it will be over multiple years. Battery duration is 2 hours, depth of discharge is set to 100% and round-trip efficiency at 85%. I also created a few other scenarios where I consider:

• Adding an EV with a 50kWh battery, 7kW charger and all with V2G capability.
• Including some load flex of my aircon to see how that might reduce the battery size I need
• Including a small biofuel powered generator as an alternative supply source for the few hours each year when the air con is cranking. That one’s just for fun.

We’ll talk more about those other scenarios in another post but as a bit of a teaser I will say that the EV alone (so no additional battery) reduced my grid consumption by nearly 50% compared to BAU, despite increasing my overall energy requirement, and also that my constraint moved from being a summer storage capacity issue to a lack of solar generation in winter.