#4 Let's Make Electricity - Part 1: Sizing the System
When someone mentions off-grid, one of the first things that come to mind is not being connected to the electricity grid so it deserves some extra attention. This is probably going to be the longest and most involved blog entry I'm writing in the entire series, and I will have to split it into two parts. The reasons for this are the overall scope of the topic, ranging from solar panels, batteries, the choices to be made and how much it costs. It can be quite an involved and very technical process. Starting with power audits and analyzing current and future demands, this usually leads to an outcome that is different for everyone.
I'm going to try and build this up from the bottom and show you the decision process involved. This process, independent of actual end result, is the same for practically any off-grid situation. In contrast to my other blog entries, I'm going to provide links to calculation tools, hardware, and so on. I'll try to present both the way I did it (which is extremely DIY oriented), and possible alternatives keeping the current evolution of things like batteries and other equipment in mind.
Location again...
Because we're at a pretty high latitude (> 60 degrees), I have to take some environmental conditions into account. Let's start with some monthly solar irradiation numbers provided by this tool.
Compared to countries more south, we have of course much less solar irradiation. Furthermore, the available solar is concentrated in the summer months, and I have to find a way to use this power when it's plentiful. All this means I have to budget for a larger array than you would in other locations. You'll have to make a similar investigation for your specific region if you want to go off-grid with solar. That said, you could probably use my situation as a worst case scenario!
The First and Second Rule
The first rule of off-grid solar is: conservation is cheaper than generation. The second rule is: find alternatives. When I started building this, I could start with a clean slate. It would be much harder to turn an existing building off-grid because the choices one makes are different. Take for example heating. Here in Finland a new building would most likely use a heat pump as primary means for heating the building. Heat pumps are great: you get more warmth per kWh than if you would by just using the kWh directly in a resistive heater. The reason is simple: moving heat takes less energy than creating it. That's what a heat pump does, and the CoP (coefficient of performance) of a heat pump indicates how many kWh of heat you get out for every kWh of electricity you put in. But here lies the problem when going off-grid: you don't have enough kWh in winter to begin with, and while you could get for example a heat pump with a CoP of 4 or more, you still need that 1 kWh of electricity before you get that 4 kWh of heat. Those same 4 kWh of heat can be generated from 1 kg of properly dried birch wood without a need for electricity. If you were located in a cooling dominated climate, having a large amount of solar electricity available when you need to run the A/C is a much better match.
This outlines one of the aspects of going off-grid, especially in a cold climate: you need alternative energy sources. You can't rely on just electricity. As an example: I can cook on induction for a big part of the year, but come winter I need an alternative. For me, that's propane gas. Yes, I could have gone with a wood burning cook top as well, but it's one of the concessions I made: cooking on gas is awesome. Since it's also only part of the year I need this, it's not very expensive either.
The first rule dictates energy efficiency. Things like a fridge, the ventilation system, lighting, pumps, washing machine,... in other words the essentials that you can't do without or for which there are no real alternatives, and that are used frequently. These have to be selected in function of efficiency. Lighting for example: LED is pretty much the only way to go. Likewise, there have been huge improvements in white goods efficiency in the past decade, so get a new appliance instead of an older model. Check the energy labels - yes, they're not ideal, but they're a very good starting point. Whatever may be the case, it is at least possible today to do this; going back a decade this project would not have been feasible because of technological limitations, or it would have been prohibitively expensive.
The Audit
So how do you find out now how much electricity you really need? For that, you have to perform an energy audit. This entails recording (measuring or using numbers from labels and datasheets) power consumption numbers and run time numbers for each electrical device you plan on running. Here is mine:
These numbers assume a couple of things. For one, these are the bare minimum needed so we as a family can live a normal life. In summer, usage will be much higher (induction cooking, wood processing, etc.), but since there is ample power, it's not a concern. You will see from these numbers also that there is no oven or micro wave oven present. The function of those is filled in by the masonry fireplace in winter - more on that in a later blog. No coffee maker? Of course: a Moka pot on the gas stove.
The lighting assumes all lights are on at the same time for 6 hours a day, so it's also a sort of worst case scenario since that doesn't happen in general. The ventilation is necessary and can not be replaced or lowered. The fridge/freezer number is based on the energy label, as is the washing machine - and these are pretty accurate actually. The sewer fan is part of the waste water processing plant and is also something that can not be improved upon. The washing machine is expected to do one load of laundry per day in this calculation: it's usually less. Well pumps have a high power requirement, but they only run for short periods of time so their energy requirement is fairly low. The other pumps are circulation pumps (e.g. underfloor heating), but these are very low power and don't run all the time. The laptops assume two people working from home, and both are high efficiency devices. The Wifi router is to provide an internet connection over cellular. Phone charging takes negligible power. The biggest thing I could improve upon are inverter losses, and I did that already - but more on that in the next blog.
You can get more details on running an energy audit, and a ready made spreadsheet from this link.
From this energy audit you can see that I will need around 5kWh per day to have a comfortable, normal life situation. I also want to take a margin into account on these numbers just to make sure, so let's assume this is closer to 6kWh. Now that I have this number, I can start calculating what this means in terms of solar panels and battery storage.
Solar Panels and Battery Calculation
One number we need is the slope, or angle, the solar panels will be mounted under. Because I planned to have the panels on a ground mount, we can be flexible with this and find the optimal angle. There are tools such as this which help with that. If you don't find your exact location, pick one nearby. Tools like this give you a couple of angles, each optimal for a particular season. While I could put them practically vertical (optimal in winter), this is a bad idea because there won't be enough sun regardless. Therefor, I chose the spring/autumn angle - this will still be more than good enough for summer anyway. I ended up with an angle of 25 degrees (or 65 degrees depending on which corner you use to measure this). These numbers also right away determine how the ground mount will look like.
For the batteries, I needed another parameter: days of autonomy. If there would be no generation because of cloudy weather, rain, etc. how many days would I need a buffer for? As many as possible of course, but I started with two to three days. Based on the 6kWh/day consumption figure, this would mean a battery between 12 and 18kWh assuming we can use the full capacity. With certain batteries (lead acid) this isn't really feasible, since you don't want to discharge more than 50% to have a good cycle life. If I were to go with lead acid, that would thus mean a capacity of twice those figures. Lithium fares much better in this field. I will assume for now that we can discharge down to 1% and work with these numbers until we select the actual battery chemistry.
Going back to a previous tool, we can start to punch in the angle, our consumption, battery capacity, play with some solar array sizes, and some other numbers and see how they stack up. This is what I initially came up with after some trial and error as a first result:
Yeah, that looks bad. That won't work at all: even in summer I would have days where the battery is empty. For reference, this was a system sized with 3kW of solar panels and 6kWh of battery. Remember I started this project many years ago, and this was what was available back then at a 'reasonable' price. In addition, energy efficient devices were less prevalent, and LED bulbs were not available, or very expensive. In other words, I had to wait for technology to catch up with my dreams. For example, back in 2008 the cost of a solar panel would come at around $4 per Watt. That would mean the above system would cost $12000 just for the panels. When I finally got my panels, I paid $0.24 per Watt - that's a factor 16 less over a decade! I won't even go into the battery cost at that time...
So this what I ended up with in the end:
That looks much better. As I said before, winter months won't work on solar no matter how many panels you have. The size of this system is 10kW of solar panels with 28kWh of battery storage. That covers three to four days of autonomy, and plenty of power to not have to worry at all for eight months out of the year. Even February isn't that difficult, so the only difficult months are November, December and January. Pretty much as predicted. The array was also sized this big to accommodate summer uses (wood processing for example, more later) and a possible EV in the near future. In addition, I did anticipate that some panels might not be optimally placed, and could experience shading at times.
I'll give a complete break-down of all components, why they were chosen and with pricing, in part two. What I can tell you now already, is that the entire DIY system (including solar panels, battery, charge controllers, inverter, etc.) cost the same as just those 3kW of solar panels did in 2008.
Comments
Post a Comment