Oh man this one is a lot longer than I expected. But it’s been simmering for like 2 months so maybe it’s time? Who cares, let’s dive in.
Connecting the Dispersed
So in this final portion of our series on Power and some of the less obvious aspects about it, we want to talk about the connection between utilization, generation, and transmission of power, more commonly known as The Grid.
When we consider power in light of our previous discussions about the idea of dispersing the generation capacity of our society and determining how we allocate this, we are drawn to one topic about The Grid in particular. This is a small portion of the overarching idea of “smart grids”, called micro-grids.
Now micro-grids aren’t actually that interesting by themselves. In fact, they aren’t even that complicated. A micro-grid is simply a small subset of the grid, usually in a local area, say everything attached to a particular distribution sub-station. The important thing about a micro-grid is that it has enough local power generation on That side of the substation that, if necessary, it can be cut-off from the larger grid and continue to operate independently for an indefinite amount of time. Great for times when a major power line gets downed in a storm or other natural event.
But the Really interesting part about Micro-grids is the question that they actually ask us.
How Small can a Grid Get?
Now before we dive into this, we need to define what we mean by a “grid”. A single house with solar panels and a battery could be the minimum, but we are talking about how micro-grids may allow us to divest ourselves of Power systems that we might not approve of. Just saying that it’s a single house would basically be saying that everyone has the option to go live in the mountains and store their poop in bags to use for fertilizer. And you know what, if you want to do that then more power to you, but I like warm showers and indoor plumbing and I’m betting that I’m part of the majority in that.
So what I think is a good mix is that our Grid is defined as “as small as possible while still supporting the livelihoods of the people who live in said grid for the lifespan of the infrastructure that powers the grid”. When you start getting large enough to not merely maintain but completely rebuild/expand your infrastructure, then you are starting to get into the realm of a full scale grid or arcology depending on how you look at it and that’s not what we want to explore here.
Local Energy Production
Since micro-grids must be self-supporting for power requirements, we have to consider that all of the power is generated on site for the micro-grid. So how do you supply the energy needed? Well, to determine that so that no one has to deal with blackouts, we need to know two things; the population density, and the average power usage per household.
I’m feeling a bit nostalgic today so lets take the neighbourhood I grew up in as a case study for this. Greenfield, Edmonton has about 3600 people living in it, and covers an area of about 1.5 square kilometers. This gives us a population density of right around 2400 people per square kilometer, so pretty average suburbia. And the average Albertan uses approximately 2400 kWh/ year in electricity which is equivalent to an average draw of 274 Watts per person (not total energy or even household energy which is important, but we will get to that later). But in total these numbers mean that the total electrical power per area that Greenfield needs is about 0.657 W/m^2 based off of 274 W/person and the previously mentioned population density.
W/m^2 is an interesting unit for these concerns. It means that if we had a way to produce power that only produced 0.657 W/m^2 that we could provide enough energy to power all of Greenfield within the area of Greenfield. But that means that All of the area would be take up by power generation. So unless we are planning on living underneath or on top of a giant plate that houses all our energy generation, we need something a little more energy dense to make our electricity while still leaving us enough room to live, travel, and actually enjoy our community.
Another reason I’m using that unit is because it maps onto this chart from David Mackay which is extremely useful.
The chart showcases thresholds of energy density needed to provide enough energy to supply the denoted country. So long as the country dot is to the lower left of a line it is possible to provide enough energy to that country from that source of energy. But… the closer the dot is to the line, the more of the total area of that country must be devoted to that energy source. So for example, the United States can Technically provide 100% of its energy requirements from biofuels derived from energy crops, but it would take 50% or more of the Entire landmass to do it (including Alaska, which doesn’t grow crops very well so… yeah, probably means that more of the good land would have to be taken up for that).
The farther up and to the right a country is on the chart, the fewer options are available to it. The UK is too dense for biofuels, South Korea is too dense for Wind, and Singapore and Hong Kong are too dense for even Concentrated Solar Power. So what about our theoretical Greenfield Microgrid?
So as the new dot shows, Greenfield is a bit too concentrated to be able to use biofuels to support its electricity needs, but it is able to be supported by wind power. (Note: there is some contention about this value of energy per area of wind power and I feel the need to state that this is for onshore wind production as offshore turbines are a little more space efficient *On A Per Turbine Basis* due to their ridiculous size, and also because Alberta is nowhere near an ocean.) In fact, the average power requirements for Greenfield come out to just under 1 MW of electrical power, so a standard 2.5 MW wind turbine would be able to cover that requirement with only a 40% capacity factor, which is a little on the high side for onshore wind but not totally unbelievable so lets go with it.
But as we’ve mentioned before, several times, wind and solar are inconsistent in their generation and as such need back-up to provide power when people want it rather than when the weather provides it. Since the production times of wind are effectively random, it is a fair assumption to say that with a 40% capacity factor that it will be producing power at ~40% of the time that it is needed, so to provide enough energy at all times we need a battery system capable of providing 60% of the necessary energy for the community for a predetermined amount of time. How to determine that time is complex enough for a whole article by someone who is much better than I at grid management and analysis.
Lets take the easy route out and say 60% of 2 weeks worth of electricity to cover >99% of possibilities. From our previous 274 W/person that means that our battery will be 179.3 MWh to cover 60% of demand for a 2 week period. As a bit of perspective, the giant Tesla Battery in Hornsdale, Australia is only 129 MWh and cost $100M USD. I don’t think that a community league bake sale will be able to raise that much.
But lets be pragmatic, after all, wind power isn’t the only line on that previous chart, throw solar panels on all the houses and you can easily help even out the production levels to get rid of the majority of zero production hours during the day. But it still won’t prevent there from being large dearths of generation during the night because, we can’t pull in energy from farms in different parts of the country as remember our initial assumptions are that this micro-grid is able to operate independently for the lifespan of the energy infrastructure. So no using the standard get-out-of-jail-free card of pulling energy from wildly dispersed locations to counter local weather deficiencies. But a mix of solar and wind seems like it would quite comfortably supply all of our micro-grid, especially if a single windmill could do it, albeit with a $100M battery backing it up. And yeah, the numbers seem to show that it could probably reliably handle all of the electricity demand….
But what about Total Energy?
This is where things start to get less rosy. While the average Albertan uses 2400 kWh of electricity per year, we use Vastly more energy in the form of natural gas. About 110 GJ per household per year, which according to Google, works out to ~10,200 kWh per person per year. This moves us from 6 kWh/d/p to 34 kWh/d/p and ups our little micro-grid on the chart by about this much.
So 10,200 kWh/p/year for heat, 2400 for electricity, and 3600 people gives us a total yearly energy of 45.36 GWh. To provide this amount of energy at the 5 W/m^2 value for northern European equivalent solar (which is a proper comparison as Edmonton is at the same latitude as Minsk, Liverpool, Dublin and Hamburg) then it would require 1.04 square kilometres of solar panels to power the community. As Greenfield is only 1.5 square kilometres, you can guess that this is Far more than the available roof space. Even adding the 1-2 windmills (the max that can be supported on the available land area is actually 3.75 MW so only about 1.5 of the previously described windmills, but lets say we got lucky and were able to place them right on the edges of the community grounds so they interfere with each other the least. ) only cuts down the needed solar area to 835,600 square meters or 635,600 square meters respectively. Which is still nearly half the total area of the community even in the best case and significantly more than the available rooftoop space of all the buildings.
And of course this ignores the potential inefficiencies of converting electricity into heat as compared to just burning natural gas for heat, and I don’t feel qualified to even Attempt to figure out the kind of battery back-up that would be required to store that amount of energy.
So if we don’t want to fill every available non-residential space in our micro-grid with energy generation mechanisms, and also don’t want to sprawl the energy generation far outside where the power is needed, we need a way to produce power that is compact, produces a lot of heat that can be used directly, and a smaller amount of electricity in a consistent manner that is controllable and rampable.
If Only there was Something in development that met those Requirements
Now, just stating that nuclear can do that and then walking away would be a kind of douchebag, echo-chamber thing to do as it would make it look like I don’t hold nuclear to as high a standard as I do other types of energy generation. And that would make me no better than a Fossil Fuel promoter. So lets take a look at the average land use power densities of nuclear generation and do some extrapolating.
Here’s where we draw on our old friend Bruce Power Nuclear Generating Station. I’m sure our long time readers can almost recite these statistics by rote at this point so lets just breeze through the set-up nice and quick:
Bruce Power Nuclear Generating Station
land area: 932 ha (9,230,000 m^2)
power generation: 6430 MW
capacity factor: 87.4%
Area energy density: 608.86 W/m^2
609 W/m^2. That’s enough power to provide two Albertan’s average electrical requirements per square meter, and enough to need only about 3 square meters to supply more than the 1.43 kW average total residential energy usage of that same person. But, obviously the Bruce Power Plant is hugely larger than the community of Greenfield, so it’s not very practical if we are trying to keep the generation requirements inside the footprint of the micro-grid, and it certainly doesn’t lead to a very positive answer to our previous question of “how small can we make a microgrid?”
But not all civilian nuclear sources are huge hulking things. At least, they won’t be for very much longer. While we don’t have any definitive numbers on land usage of SMR facilities, there are a couple of companies that have released tantalizing details that we can use to make some educated guesses. Those companies are Thorcon, and Oklo. We’ve never mentioned Thorcon on this website before because we try to focus on how Alberta might benefit from nuclear technologies, and Thorcon’s whole thing is based on nuclear facilities located on Oil-tanker sized ships that are parked to produce power and then sail back to the main factory to be refueled when the time is right. And since Alberta is completely land-locked it doesn’t seem like it will have much direct impact on us, So lets focus on Oklo.
Now Oklo isn’t currently publicly pursuing Canadian licensing but they are the only one that seems to have any information about footprint size so we take what we can get. And speaking of footprint, Oklo says their plant will produce 1.5 MW of electricity and be about this big.
So a 5000 sq ft building (464 m^2 for us upstanding metric users) plus up to 1 acre land to make it look pretty on the outside. Earlier statements said that it would be smaller than 1 quarter of an acre (<1011 m^2). This gives the Oklo reactor a land area *electric* power density of *drumroll* 3232.7 W/m^2 for just the building, 1483 W/m^2 for the smaller land plot, and 370.7 W/m^2 for the larger plot!
That’s a pretty big range, going from about half as dense as the Bruce Power facility to almost 5 times as dense depending on how removed from the community you place the facility. This is how we figure out how to make the smallest microgrid. Given that it isn’t a direct electricity producer and converts it from heat, we can assume that the reactor produces about 4-4.5 MW of heat with about a 30-35% conversion rate to electricity (The Oklo design uses heat pipes rather than directly heating water so it is unclear as to if it uses a steam turbine or a Stirling engine to make electricity, but with the small footprint I’m betting on Stirling engine). What this means is that one reactor would be able to supply 50% more electricity than everyone in the community uses, and a second one would be able to produce all the Heat that everyone uses if there was a district heating system. And since, unlike solar panels or windmills, the rate at which this energy can be produced is Controllable, you wouldn’t need any extra infrastructure such as batteries to act as support for the grid. Thus with the minimum amount of moving parts, the whole community could operate pretty much completely isolated from the national grid for the 20 year lifespan of the reactor fuel.
But remember, these figures are only for a single, fairly average, and kinda sedate neighbourhood. If your goal for a micro-grid is more self sufficiency, then you either need to increase the amount of land available for things like farming, recycling, composting, and other forms of employment, or increase the amount of energy that people use. You want to shrink the amount of space that people use, you build up. High density housing, walkable neighbourhoods, vertical farms. These are all ways to reduce the amount of land that humans use, to allow more of it to return to nature. Unfortunately these methods actually increase energy usage overall even if apartment living is less individually energy intensive than suburban living.
Even so, self-sufficient micro-grids powered by small nuclear reactors take significant strides toward the goal we started this whole series with. They could allow small communities to achieve some measure of autonomy and self sufficiency while requiring less investment in infrastructure than other options. They would allow the grid to become somewhat more dispersed and therefore resilient, but still leave room for people to actually enjoy the space they live in. All while reducing our dependence on fossil fuels.
That’s the sort of Power worth pursuing.