Circular reasoning (Rolfe's solar energy system)

[Gomev]

You need to treat yourself to one of these...

https://www.argos.co.uk/product/4821261

Only draws 1kW

It still takes the same amount of energy to heat a set amount of water whether you draw 1kW or more. It just takes longer. So lower draw is useful if the output is restricted, but it does not actually save anything.

 

[Gadget Geek]

I have a Vektra thermos kettle. You can set the temperature and it keeps the water hot for when you are ready for another cuppa. If it's cooled down too much, it's just a quick reboil as the water is still relatively hot.

Best kettle I've ever had.

 

[GaryMG4]

It still takes the same amount of energy to heat a set amount of water whether you draw 1kW or more. It just takes longer. So lower draw useful if output is restricted but does not actually save anything.

I think the idea is that if you have 1.2kW excess that is being clipped then there is no cost as you don't use the grid and you don't get paid for that clipped excess.

 

[Rolfe]

It's not quite like that. In order to harvest the generation that would otherwise be clipped I have to leave the inverter set to export the battery down to 100%. Which is very counterintuitive, but that's how it works. That way the battery only charges when the generation is over about 5.3 kW. Anything below that is exported. This is important because without this setting the battery would charge from the solar preferentially, get to 100% pretty quickly, and then I'd lose the clipped generation.

When the battery is on this setting it won't discharge to power the house. If there is enough solar to boil the kettle then that's fine, it will use that rather than exporting it. But if there's a cloud over the sun at that precise moment, so that the solar isn't enough to meet the kettle load, power will be taken from the grid. At the daytime rate.

So Everest is quite right. A kettle with a lower power setting would solve this by bringing the kettle draw down to what the solar can cover. I'm just too impatient. I want my coffee  now. It's quite easy just to turn off the export setting, boil the kettle, and turn it back on again.

 

[T1 Terry]

It still takes the same amount of energy to heat a set amount of water whether you draw 1kW or more. It just takes longer. So lower draw useful if output is restricted but does not actually save anything.

Not actually correct, unless you have a device that is so well insulated there is zero heat loss.
An example, those cigarette lighter elements that you drop into the cup to boil up the water for a cuppa, really struggle to reach boiling point even if the inside of the car is warm, leave it outside in winter and the water will never boil, the heat loss is greater than the thermal input and the closer to boiling point, the greater the differential between the heated water and the outside temp, so the greater the heat loss .... I used these a lot when balancing big battery packs during that initial conditioning charge required for long life when used as house batteries

Another example, more than once I've had to create a blanket around a slow cooker so it would actually cook, and it isn't that cold over here

T1 Terry

 

[Everest]

It still takes the same amount of energy to heat a set amount of water whether you draw 1kW or more. It just takes longer. So lower draw useful if output is restricted but does not actually save anything

Technically a lower power kettle will actually use more energy to heat the same amount of water due to heat losses during the heating process. But the point was about power, not energy demands.

So Everest is quite right. A kettle with a lower power setting would solve this by bringing the kettle draw down to what the solar can cover. I'm just too impatient. I want my coffee  now. It's quite easy just to turn off the export setting, boil the kettle, and turn it back on again.

We used to turn things (like dishwasher) off when we had only a 3.6kW inverter (a few years ago) if we wanted to boil a kettle... but got fed up doing that - hence used a low power kettle for a while. But I agree it's a pain with the time it takes to boil!

 

[GaryMG4]

I want my coffee  now. It's quite easy just to turn off the export setting, boil the kettle, and turn it back on again.

My coffee machine uses about 1.3 kW and only heats enough for the cup so I'm sorted. Mind you the price of the coffee pods far exceeds the energy 'saving'

 

[Rolfe]

There's more cloud again today and even the morning had high light cloud, but the battery has already put on 26% (from 39% to 65%) and it's charging at over 2 kW whenever the sun comes out.

 

[Rolfe]

It worked properly today. It's worth posting the graph. I've taken out the grid import because it was messing with the scale, but that doesn't matter so much.

At midnight the home battery is charging from the grid. The dishwasher is also on and is running at the same time, also drawing from the grid. Once the home battery is full, the Eddi heats the water for up to an hour, while the battery is held at 100% by still being on charge so again the Eddi draws from the grid. After that the battery discharges to export for an hour and a half, bringing its SOC down to 39%. (If I want it less than that I need to schedule a slightly longer discharge window.)

That's the end of the off-peak tariff. The battery is now free to supply the house until the solar takes over, but it only dropped by 1% for about ten minutes doing that, because there's not much house load at that time and the solar was beginning to ramp up.

Sunrise, and the solar quickly takes over the house load and starts to export. The battery can't take the solar because it's set to export to 100% (yes, I know) from sunrise. There was some high cloud so generation wasn't what it might have been, but by about ten o'clock there was enough for the export to be clipped. The battery then took the excess over and above that, and did pretty well until lunch-time when larger cumulus clouds started to come over. I ran the microwave at lunchtime without thinking about it, but even though generation was in a trough there was still enough to cover it.

The afternoon was cloudier, but direct sun during three or four breaks in the cloud let the battery get right up to 78%. Overall it harvested 3.9%, which is about 4 kWh. By about 3.30 pm when I made coffee the generation had fallen off a lot, and was at another nadir. I flipped the battery export off as I turned the kettle on, then back on again when it had boiled, just to stop the kettle drawing from the grid. Then the export continued.

There was still a bit of export by six o'clock when the export slot finished and the battery put on a couple of per cent, with a slight wobble as I boiled the kettle again. As the solar faded the battery took up the house load, starting to discharge at about eight. Having looked at what was left I set the battery to start exporting again just before ten, so that it finished discharging at about 11.30. Back to the beginning again, except the forecast for tomorrow is so bad that I'm not going to bother discharging the battery in the early morning as I don't think there will be enough clipping to make it worthwhile.

The last green peak of load is just because the car started to charge. I had plugged it in but then decided not to as it was already at 67%, so I turned it back off again.

So this works, it's only a minor inconvenience to keep an eye on what's happening during the day, and even on a day where there wasn't a lot of direct sun it gained enough to be worthwhile. I exported 10.3 kWh from the battery as well as 31.8 kWh solar generation. (The battery is only 9.5 kWh, so doing it in two stages like this with some charging in between definitely paid off.)

 

[Gomev]

Not actually correct, unless you have a device that is so well insulated there is zero heat loss.
An example, those cigarette lighter elements that you drop into the cup to boil up the water for a cuppa, really struggle to reach boiling point even if the inside of the car is warm, leave it outside in winter and the water will never boil, the heat loss is greater than the thermal input and the closer to boiling point, the greater the differential between the heated water and the outside temp, so the greater the heat loss .... I used these a lot when balancing big battery packs during that initial conditioning charge required for long life when used as house batteries

Another example, more than once I've had to create a blanket around a slow cooker so it would actually cook, and it isn't that cold over here

T1 Terry

Not to mention of course that the day has an T in it, you are wearing a woolly hat and wellies and it's between 2 and 3pm.

 

[Rolfe]

I've been doing some more experimentation, and I've found a better way to work things which is also very easy to operate. Today I seem to have it working perfectly. Here's the picture that tells the story.

Start charging the battery at midnight (having exported whatever was left, to end at 11.30 when the peak rate ends). It reaches 100% at 03:15. However I keep in on charge for the whole of the off-peak tariff to force the house to use the grid. Not a lot happens until 04:30 when the Eddi comes on to heat the water. It doesn't always take the full hour, but I give it till 05:30 anyway. That's the end of the off-peak period, when the water-heating stops and the battery is no longer on charge, so it can start powering the house load.

So far, this is just the same as the winter configuration. However, if clipping seems likely, I need to get the battery down low enough to accept that. So I start to export the battery at 05:30, stopping at 07:15, which gets it down to 30%. This is plenty to run the house for the evening in the event that the forecast is 180° wrong, but still leaves enough space to take any likely clipping (I think).

This is the main change from before. Yes the battery is still exporting after sunrise, but at that time the amount of solar, added to the battery export, isn't enough to push the total above my export limit. It also means the early morning is covered with no possibility of grid import if it's dull. (I may have to move this time back a bit as the sunrise gets earlier, we'll see.

At 07:15 the battery is now set to export to 100%, the setting that holds it at the SOC it had at the start, allowing it only to take the surplus generation that might otherwise be clipped. This holds till 7 pm, when the solar is fading and the battery begins to be needed to run the house. In practice though, if I start to cook earlier, I'll just switch the export setting off at that point.

The battery runs the house through the evening, then I have to decide when to start exporting what's left to finish around 11.30. This is the bit I can't really automate it seems.

This has the advantage that it's mostly the same as the winter settings and I don't have to change much. I can charge the car between 11.30 and 5.30 as usual, while the battery is on charge, the water is heated last thing before the end of the off-peak tariff either way, and it doesn't interfere with washing machine or dishwasher use. All I need to do in the evening, when the battery has finished exporting, is to decide whether to leave the export slots active, so it does its clipping-harvesting thing, or switch them off, so that it goes into the morning at 100% and is always available to the house if required. That's where the weather forecast comes in. Any wee sunbeams showing in the middle of the day and it's probably worth it.

On the graph above you can see the battery charge from the grid from midnight until 03:15, then it's all quiet till the water heats 04:30 to about 05:15.
At 05:30 the grid stops powering the house and the battery starts to export. Towards the end of the export period the solar export is added to it, but in total it never reaches the 5 kw export limit.
At 7.15 the battery stops exporting and holds its 30% charge while all the solar goes to export. There was a bit of high, hazy cloud, but by about ten the generation had reached the clipping limit and the battery began to charge. It made it to 66% by about one o'clock, but after that there was more cloud and it didn't charge much more. The shenanigans about three o'clock were me boiling the kettle for coffee, having switched the battery export setting off momentarily just in case. I didn't have to do anything while I was making lunch as there was plenty solar to cover it.
I didn't start to make the tea until just after seven so I didn't need to alter anything. After that the battery was free to power the house (and get any solar that was still generating).
You can see where I turned up the heating about 8.30, and then at 10.40 the battery started to export what was left, as I had set that time some time earlier in the evening. I think it finished about 11.32, and then the grid took over, ready to start again tomorrow.

Basically a perfect demonstration, and I'm printing this one.

 

[Everest]

Well done.... one really gets a feel for solar production and house usage over time.

Here's the picture that tells the story.

My reading of that is:-

• @Rolfe has a very low quiescent house load.
• She only has one cup of coffee in the afternoon and no tea, apparently
• Her evening meal was heated in the microwave (guessing being a load of about 1500W) at about 19:30
• She didn't do any washing or tumbling yesterday.
• She didn't use a dishwasher.
• But how did she heat up her porridge?

 

[Rolfe]

Well done.... one really gets a feel for solar production and house usage over time.


My reading of that is:-

• @Rolfe has a very low quiescent house load.
• She only has one cup of coffee in the afternoon and no tea, apparently
• Her evening meal was heated in the microwave (guessing being a load of about 1500W) at about 19:30
• She didn't do any washing or tumbling yesterday.
• She didn't use a dishwasher.
• But how did she heat up her porridge?

I did go round the house at one point checking how I could minimise the base load. The two things which mainly boost it are the central heating boiler running, and my hifi system which often runs most of the day. I think the central heating didn't really come on yesterday because of the sunny weather, until you can see where I turned the temperature up just before nine as it was getting a bit chilly. The additional drain on the battery after that is obvious. Looking closely I can see where I turned the sound system off after lunch and went out, and turned it on again when I got in and made that coffee.

I don't usually have porridge if the weather is warm. If I did have it, it would be on the smallest hotplate of the cooker, not really sure what that draws, and it only takes a few minutes. I didn't heat anything for breakfast. I did microwave lunch, but the graph (which only records every five minutes) didn't happen to pick it up. I must have boiled the kettle for the coffee just on a five-minute mark for it to be so obvious. I microwaved at teatime too, and boiled the kettle, but again the five-minute point missed most of it. The microwave is rated as 800 watts, not sure what it draws, I suspect less than 1.5 kw,

The washing machine and the dishwasher are run after 11.30 to draw 7p/unit mains electricity rather than solar/battery which could be exported for 15p/unit. The reason none of that happened is that Octopus had a deal going the previous night where if you charged your car (but didn't charge it the following night) the price went down to 3.5p/kwh. So here is the previous day's graph. (I've taken out the grid import as it was messing with the scale.)

You can see the massive overnight load, which was car charging and running both the dishwasher and the washing machine. I seldom use my tumble dryer, I just hang the washing up.

That was my final trial on the system as set up. I decided that I didn't want the battery to go as low as 20%, and that 7.30 am was too late to still be exporting anyway (it almost started clipping as the solar had ramped up), so turned it back to 7.15 to get 30% SOC remaining at the start of the day.

 

[Rolfe]

Today was another object lesson. When I woke up soon after eight everything was fine, battery at 30%, solar ramping up. But by nine, when the weather forecast had a nice round sunny face on it, what was actually happening was black clouds and heavy rain. The generation fell to 100-150 watts, barely enough to cover the house load. It was covering it, but I thought, if the central heating decides to turn on, it will draw from the grid. So I switched the export slots off, and of course soon afterwards the clouds cleared and the battery started to charge from the solar, so I switched it all back on again. It's very mixed weather. There's no real danger of drawing grid power in the middle of the day unless I start to use the vacuum cleaner or something, but one minute it's sunny and the next it's raining. The battery got another 7% by harvesting clipping, but now generation is back down under 1 kw.

It's OK, I can ignore it, and even if it had swiped some grid power earlier in the morning it would have been very little. But it is a warning not to start hoovering or cooking if it's dull without turning the inverter settings off. I'm going out anyway. When I'm out and the house is only on base load, it's fine.

 

[Everest]

The microwave is rated as 800 watts, not sure what it draws, I suspect less than 1.5 kw,

You'd be surprised... we were when we started measuring things. Most microwaves are only about 50 to 60% efficient and the power relates to energy delivered to the food. Our 900W one draws just over 1500W.