I’m currently in the throes of building a 1.2MW solar farm somewhere in England. Its an adventure, to put it mildly :D. So far the planning process and bureaucracy has taken just under 2 years. In that time a lot has changed.
Originally the plan was for a 1.2mwp solar farm, and a battery of either 256kwh or 512kwh. The current situation is that I’m evaluating a 583kwh battery (they can basically pack a bunch more cells in there now), but the economics have got a bit trickier and it may not actually be viable. This does not mean the project itself is not viable, the farm likely still is profitable (I hope!).
First things first, I’ll explain why any of this would make any sense. You cannot just rent a field, fill it with solar panels (assuming the farmer lets you, and planning permission is granted), and make a pile of money. Its not that simple. (And trust me: none of that is simple). The main problem you face is the grid connection. Basically solar farms go in fields, so we are already limited to relatively rural locations. This means that the land is large enough, and also its not going to upset too many people. Our site is surrounded by fields and hills and generally…nothing, so there were zero objections. This is all good news..
…but it also means you are on the edges of the power grid. If you are REALLY lucky, you find a site that has a field to rent, thats rural, but also just-so-happens to be a site where a major grid connection cable runs right through it. This is super ideal, and no, there no sites left like this any more in the UK because all the canny early developers have already snapped them up :(
So…we end up with a perfect field, planning permission, and a happy landowner, but we are basically connected to the rest of the grid by what amounts to a thin USB cable (not really). This is the point at which you ask the grid to upgrade that line and send you the bill, they do, then you laugh at how they accidentally added a few zeroes to the bill, and then you start to cry. Basically upgrading a power line is a nightmare. They cant just add heavier cable and be done. That cable may then sag too much in summer, or be too heavy, so you need new pylons, and then the ground under the pylons is too weak so you need foundations and omg expenseahontas.
In practice, that means that the local DNO (power company) has told us ‘you can generate 900kw at any moment from here but no more’, and thats AFTER we pay for a major grid upgrade. This sounds nuts, but when you think about it, if a power cable was only ever built to supply 2 farm houses, its not ever expecting a power flow of more than about 40kw total maximum. Then suddenly we arrive on the scene with 3,000 solar panels and want to push 20x as much power through those cables…
What does this have to do with batteries?
You might have noticed the difference between our 1,200kw solar farm and the 900kw limit on export. Well spotted. Its quite glaring isn’t it? Actually, this is exactly what most people do, and if you have solar panels at home, you do it too. Its called inverter under sizing.
The inverter is the box that converts DC to AC, and is the device that determines the maximum power that can come from your solar panels. In many cases, its rated BELOW the maximum output of your panels. This must seem insane…because surely that means at peak sunshine some of your power is just wasted right? YES. This is exactly what happens, and its normally because the bigger inverters are very expensive, and the PEAK output of your solar panels only triggers quite rarely, maybe for just 1 or 2 weeks in summer around midday. This means that the extra spend on the ‘bigger’ inverter might not be worth it.
Example: You have a choice of a 2kw inverter or 2.5kw inverter. The difference is $1,000. You have 2.5kwp of solar panels. Which do you choose? Well its actually hellishly complex. You need to model the output of your panels for every hour for every day in a typical year, and then ‘clip off’ the top portion where output would have been above 2kw. You then need to value that power at whatever you get paid to sell it, or what it would have cost you to buy it. You do that for the lifetime of the inverter, then work out if its a good deal.
I’m currently basically doing a similar thing, except the limitation for me is export limits, not inverter sizing. Our site will have 10 inverters anyway, and the bigger you get, the more granular you can be with how many you have. (For those curious, yes, thats like 300 panels per inverter, but each inverter supports a bunch of strings).
What does this have to do with batteries?
Well batteries let you cheat the system! They can do it in home installs, with what they call DC-coupled batteries, and you can do it at the grid/utility scale with batteries the size of shipping containers. (I’m evaluating a 20ft long one). Before I get into this, I need to list the different ways the battery might make sense in our solar farm. The battery would generate revenue basically 3 ways:
Frequency support services:
This is where the national grid basically pays you to act as a sort of ‘cache’ or ‘buffer’ for the grid. If they suddenly have slightly too much power, then dump it on you, then grab it back a bit later (maybe even a few seconds or minutes). They do this to keep the electricity grid at 50hz, to prevent every transformer in the country going bananas and maybe even failing. This is a constant battle, fought every second of every day. Here is a chart of the last 36 hours of it:
You can make money from these services, just by having a battery connected to the grid, and being registered to rent out your battery to act as a local buffer.
Time-based arbitrage stuff
You might think the price of electricity is fixed for your current contract term. Ha. Yes, at the consumer level it is, but behind the scenes, the wholesale price of electricity fluctuates like crazy, not only each month, or each day, but each 15 minutes. And it fluctuates a LOT. You can decide to ‘keep’ some of that solar power you generated at midday and sell it later. In fact in theory this can be hugely profitable, based on the current volatility. Again, here is a price chart the last 36 hours:
In that chart, the price gyrated between negative £70 (They pay YOU to use power) to positive £200 for a MWH. In theory, if you could buffer the whole farms output, and only sell it at the peaks, you would make decent money
Peak-Shaving
This is a system where you take that extra power that you are not allowed to export, you buffer it in the battery, and trickle it out later. This transforms the usual bell curve of solar output into a different shape:
We fill the battery with the ‘red’ energy, then trickle it out later to extend the peak output duration of the site. This energy would otherwise be totally lost, and we would not earn a penny for it!
Ok, so these are three ways storage can make money alongside your solar farm. What the problem? The problem is two-fold. Firstly, there has been a HUGE increase in the amount of grid-connected storage in the UK, which means lots of batteries bidding for the same grid services regarding frequency and price arbitrage. That means that naturally, the profits from these mechanisms have dropped. There are a lot of VERY big companies doing this now, and tons more on the way. Plus, there are other technologies that you compete against. The main one would be hydro (basically a gravity-water battery), but also there are experimental things like gravitricity, and companies working with flywheels and other tech. Its VERY volatile.
Secondly, the price of the battery has gone fucking nuts. In 2 years, the price has gone up 50%. Thats crazy. This may partly be due to the EV revolution, which is scooping up all those lithium ion batteries I was hoping to pack my shipping container with. It might also be related to components shortages for power electronics (lots of chips in a stationary battery), and general inflation for stuff like shipping and labour to install a concrete base and so on…
So right now, with the first two potential battery income streams kinda poor, we have to think whether the basic and most obvious one (peak shaving) is actually worth it. It looks right now that a 583kwh battery cost is about a quarter the total cost of the entire project.
That MIGHT mean its just not worth doing. Sure, being curtailed at peak output for a few weeks each summer may suck, but the big question is how much? I don’t have access to the data yet, and do not have software that accurately measures it, nor is it likely worth the hassle of buying any. I am digging further.
The maths I need is pretty simple, IF you have the modeled data: Given the output each day, for 25 years, what amount of KWH is curtailed on the site given a 900kw power limitation. I then need to guess the market price for this and compare it to the battery install and maintenance cost for 25 years. (I know, its optimistic to expect the electronics to last 25 years, but hey).
Hopefully it DOES make a convincing case, because if not, I basically have to scrap the idea of a battery, or at least wait and see if prices fall. I can still install the solar farm in the meantime.