The next power plant is on the roof and in the basement
newyorker.com
newyorker.com
Even in this short space of time I'm interested to see how I'm starting to think more about -when- energy is used. When to do daily chores (washing machine, dryer, dishwasher), when to use (limited) battery power, and so on.
I'm also tracking "discarded production", if I have enough of that then it pays me to send back to the grid. (But there are setup costs in that.)
Once you have skin in the game, you learn the rules and play a bit harder.
All you need is variable pricing, real-time load measure, and a phone app.
Also, I expect that night rates will end up being higher than day rates (if not already). Which is opposite to historical trends.
The major loads such as heating and cooling people Also want them immediately and not later.
> Load flexibility is the giant issue nobody is talking about.
Economics 101 assumes some demand elasticity, but the wholesale markets are generally fully inelastic (maybe a tiny percentage point of that is elastic in many regions, while it's finally growing in others). These markets have to use administrative solutions to approximate the value to demand as they don't have a true aggregate demand curve built out of thousands of customers that are fully responsive to price.
We're retiring our thermal fleet at a rapid pace and load curtailment will be the only option on a fully renewable grid once the renewables drop in output and the storage runs out. We are making strides in that direction.
Maybe when he says "nobody" he's referring to the general public? I know this is an extremely common topic at the utility, ISO, FERC, academic, and so on level. As the comment points out, flexibility in demand allows for all kinds of neat tricks like pre-heating things when prices are cheap and supply is plentiful. The grid is so cool.
The thermal is retiring crazy fast. NERC is very alarmed and there have been many Senate and FERC hearings on this lately. Once it's gone, it's not easy to get setup again. A lot of plant maintenance lives in the skills of the workers, so once they've relocated it can be difficult to restart. I agree with the goals of a transition, but it needs to be managed carefully and orderly so we don't drastically increase our amounts of blackouts. At the same time, we'll need to make greater strides with flexible loads of course.
So, you get energy companies that are able to adapt very rapidly to peaks and dips in energy production and demand. And that in turn incentivizes large consumers of energy to invest in batteries. Top them up when energy is cheap, sell off energy when it is expensive and electricity companies are buying. Use solar panels to get some free energy when it's available. Store any excess, and sell it when energy prices peak. Buy energy when you need to. That kind of is the game now.
If you think about the insane amounts of batteries that are going to be produced in the next few decades to power ships, trucks, planes, cars, grid batteries, domestic batteries, etc. it becomes clear that managing the charge state of these batteries is an enormous opportunity. When do you charge? Can you sell energy back to the grid when its scarce? We are talking many twh of batteries produced per year accumulating to a sizable percentage of the world's energy production (around 25000 twh). Each of those batteries can be cycled thousands of times. It starts adding up. A few decades from now, we'll have tens to hundreds of twh of battery out there. Most of them at a high percentage of charge most of the time, mostly plugged in and ready to charge/discharge in milliseconds. That's a lot of energy just sitting there. Even a small percentage of that is a huge amount of power.
You will also learn about the realities of solar vs. the fantasy of solar (as I did once I installed my system).
Like this reality (my system), cause by clouds and rain:
https://i.imgur.com/bo0s7b0.png
Most people think solar is a nice flat line of constant power once the sun is up. Not even close.
Here's another example of a day that was close to ideal vs. two day with clouds rolling by:
For the next 6 months I'll gather data to answer the question - should I have more battery (expensive cost, short-term storage, 95% effecient), or rather pump back to the grid (cheap equipment, long-term storage, 40% efficient buy/sell margins).
The answer to this question will surprise you as well. The answer is: Lots more. Well, this, assuming you want to store enough energy to be reasonably self-sufficient during weather events.
The shape of a perfect solar energy day approximates an inverted parabola. The integral of an inverted parabola is the energy you produce. This turns out to be 2/3 of the enclosing rectangle. The height of that enclosing rectangle would be the constant peak power output (think of a nuclear power plant, which is constant power).
Not too long ago, we had a period of about four months of fairly consistent mid to heavy rain. That meant heavy cloud cover even when it wasn't raining. If you want to plan for such events and perhaps, assume you might have to deal with the grid going down due to emergencies, you'll need a massive amount of batteries to store the energy. If you depend on an electric car for transportation it's even worse.
I installed a 13 kW system. That was nearly five years ago. In other words, I have years of looking at this data and navigating various hypothetical and real scenarios. My conclusion is that I would need to triple my system and have somewhere beyond 300 kWh of storage if I wanted to be self-sufficient enough for a two week event without grid power, particularly with electric vehicles.
I'm in CA, so a two week outage, while definitely black swan, is just one earthquake away. During the Northridge quake (I lived less than a mile from the epicenter then) we were without power for three days.
Anyhow, I like having solar. I just wish the people pushing it as the solution to the problems of the universe would understand the difference between solar reality and fantasy. Most of my neighbors were sold these small arrays (3 to 5 kW), have expensive leases and are in absolute pain right now as rates have more than doubled and the solar array turns out to be far from an idealized power source.
Yes and no. It gives you an idea of average production. What it does not tell you at all is the day-to-day reality, which can be dramatic (check out the screen shots I posted in my comment).
I used PVWatts and other tools extensively while designing my system many years ago. They are helpful, yet, again, none of them expose you to the reality of living with solar. In statistics, it is well known that applying averages to the entire population is a bad idea. Any single day, you can have one cloud kill half your system for hours. You'd be surprise to learn how often clouds --not even rain-- cause you to have to buy energy at the most expensive rate of the day. You quickly learn to hate clouds.
I think this is a very positive development because I my case I only get reminded of my power consumptions once a year when my supplier asks me to sent him the current value of my electric meter.
Major blackouts don't really happen, I haven't experienced any in my lifetime yet. I imagine state does have some political pressure to apply limits if something truly ruinous would happen.
Ideally people who aren't well positioned for solar (maybe there's a big tree in their neighbor's yard or something) could instead invest in batteries or other storage mechanisms. Enough people buying low and selling high and you'd see a dampening of those swings. Unfortunately panels are a no brainer and batteries are kind of a pain, so market dynamics of this sort might not help on their own.
My solar company disappeared with the job 95% done, so I can look at my bill and see that it's working, but only day-at-a-time, not hour-at-a-time.
To get better data, I can route an ethernet cable to the inverter (or buy a wifi module, which they're suppose to come with, but which didn't get installed). I'll do that eventually, but for now I'm more or less ignorant about what my system is doing. And it's been so far from my mind. Other people with panels want to geek out about it and I'm like: meh, I have different things to think about.
It makes me wonder if there are cheaper opportunities to tap into the psychology of energy usage. Sure, it would be great if everybody produced as much as they consume, but until we get there maybe we need like... outlet with little histograms or something.
I've had my 6.7kW grid-tied array for about 3.5 years, long enough to be fairly confident that it generates between 90 and 95% of our annual electrical needs.
However, we heat with electricity (heat pumps), and so our array generates 3x what we need in the non-heating season, and 1/3rd of what we need during heating season.
In the first year, I paid very close attention to stats about the system. Now, I know if makes no difference, really. It doesn't matter when I run appliances, when I recharge batteries etc. Nothing is going to change the basic facts I've just mentioned.
HOWEVER ... jump ahead a little into the future. It is possible that it may make a lot more difference to a smart grid when and how I feed the excess generated power and suck the extra power I need, and they may well want more control over that. For example (as described in TFA), they may want me to have batteries on-site to hold the daytime excess and let it dribble back to the grid overnight. Or the opposite, because they have their own commercial scale storage. It's not hard to envisage a future that moves from "I (the owner) doesn't care much about timing" to "We (the grid) care a lot about timing".
Obviously, if you generate more than 100% of your needs, you're going to want to get paid.
However, despite being convenient for the user, it isn't great for the grid nor scalable if all users were to have solar panels. (Here in Italy solar on rooftops are quite widespread).
You can see that a kWh on a sunny summer noon when all panels the area are working well, isn't worth the same as a kWh in the cold winter night, where non-renewables often supply the grid.
Often by a huge margin. Electricity prices on the spot market can go negative, and that will happen more and more often in the future on sunny, windy days. Net metering at constant $/kWh prices is unsustainable, economically.
In Britain there has been an electricity tariff called "Economy 7" since 1978. Overnight when consumption is lower, there are 7 hours with much cheaper electricity.
My parents were so used to setting the timer on the dishwasher, washing machine etc to start at 4am that it took a while for them to run them instead during the sunniest part of the day, once they had solar panels.
I have hourly pricing according to the rates on [2], but my flexible consumption is so low I don't pay attention to the price.
The same thing happens when you start using market rate electricity instead of flat rate.
You start having all kinds of apps and displays showing how the price fluctuates during the next 24 hours and adjust your usage to it.
That combined with rooftop solar and a moderately sized in-house battery is the future IMO. Charge the battery when it's cheap, use it when prices go up - or send it back to the network for a profit.
100% convenience now == burning coal and natural gas, using a gas car by myself.
95% convenience - timing washer and drier, showers (some of the time), vacuuming. - could use mostly solar + battery and no longer burn coal (SoCal).
80% convenience - time cooking - no longer burn natural gas.
Invest in a hybrid - 80% convenience because of higher price - cut about 1/3 of gas use.
Some slight inconveniences could have higher impact than others.
Wondering - Is there a resource for the impact of different changes?
I'm clocking 2 1/2 years experience with my 12.5 Kw solar glass and 27 Kw of powerwalls. Multiple electric vehicles including all yard tools, pure torque without any power loss, I have never in my life cut the grass so fast! Look ma, no gas!
Your closing line absolutely nails my thinking which is supported by good ol Lorde Kelvin, one can only improve what one measures and an extreme few know how much electricity they use. Try it yourself, ask anyone you know how much electricity they use monthly and 99 to 1 they answer in a monetary number.
Education solves ignorance and the mountain ahead is massive.
Stay Healthy!
Most rooftop solar installations were installed before home batteries and smart inverters were common, so the market often finds itself in negative pricing during the day, incentivising investment into grid-scale batteries and pumped hydro which can be paid for both storing the excess during the day, and exporting it in the evening peak. Recently, as home batteries and smart inverters have started to become more accessible for home-owners, it is common to join your household onto a 'virtual power plant' with your electricity retailer, who then commands your household battery alongside thousands of others as similar to this article and pays a credit onto your bill. The other interesting development has been the wholesale pricing retailer; not only do they apply the live wholesale price to your consumption, but also your feed in. I know one of them (Amber) can connect into your battery and drive it's charging/discharging in response to the market price. It's not been uncommon to hear people's bills ending up in credit as the capacity of their battery, combined with the price volatility far outweighs the price impact from their consumption.
The volatility poses its own challenges however. Traditional retailers which typically provide flat or peak/off-peak rate have been struggling to economically provide competitive feed-in tariffs to their customers with older roof-top solar systems; systems which are unaware of the pricing environment. Many of their customers have these solar systems which export straight into negative pricing, creating a loss for the retailer that they have to somehow recoup. As far as I understand, they can't legally have a negative solar feed-in tariff, so I assume they're shifting the loss to the consumption tariff. Even though our daytime prices are often below $0/MWh, the average Australian will tell you that their electricity bill is the most expensive it's ever been.
It works the same in the Nordics, you can purchase and sell power at the spot rate.
Our timing is still hourly, but will soon change to 15 minute intervals.
Our power here is still so cheap, and while the pricing is volatile the jumps aren't yet really large enough for battery systems to give positive ROI. At least not for consumers.
On the other hand we had literal negative prices for long periods of time during the summer and early fall so...
The last line of your comment is the most interesting to me, electricity is more expensive than ever in Australia.
From quick googling I learnt that the current home battery lifetime is 5-15 years. Replacing a battery in every home every 5 to 15 years seems like a big unnecessary hustle.
Wouldn't it be better to install all the batteries in one place, maintain them properly there and rent them out to individual home owners in the neighborhood?
I generally think the same was as you do, but for solar installs. While I think it's great that people choose to add solar to their homes, I think society shouldn't be subsidizing it until we've done more to cover commercial rooftops and other, more maintainable installation sites.
I wouldn't bet on them being that common of a battery in the future, but if I wanted a battery that would last me the rest of my life and I had the money to put down for it, I would probably put my money on an iron nickel battery.
This is not the source document, but a summary and it includes the relevant passage: https://kbelectricpa.com/emergency-disconnect-requirements-f...
https://en.wikipedia.org/wiki/Duck_curve#Mitigation_strategi...
Jokes aside, how much batteries do you need? A Tesla Model 3 has 60kWh. That's a lot of battery. For my residential needs (I have solar panels) I would be looking at maybe 10kWh.
- Is your goal battery backup or load shifting?
- How much backup do you need?
- What is your energy usage and peak demand during that time?
Peak demand is a big factor: a single Tesla Powerwall 2 for example has 13.5 kWh of capacity but only 5kW of continuous power. A heat pump or AC can exceed that, which is why many people stack multiple batteries to be able to power their homes. (The new Powerwall 3 can supply more power.)
I get solar, technically. I don't get the business of it ...