Jump to content

Solar panels without batteries


Recommended Posts

4.9 kw panels on the roof, how and when will I use any power generated ?
we both work from home. 
Highly insulated icf house, very airtight. 
ASHP UFH. Everywhere 
This is coming from my complete panic of having battery storage in the attached garage. 
I believe it’s something serious that building regs are looking into due to the increased number of fires caused by batteries, not necessarily solar batteries, but battery fires in general. 

I might be over panicking, but if I can run my house fairly efficiently then I would rather forget battery storage. 


yours mr over panicking from Cirencester. 

Link to comment
Share on other sites

Use all your big appliances, WM, TD, Dishwasher etc close to mid day, preferably one at a time.

 

Time your ASHP to only do it's DHW heating after 11AM, the sun should be up by then and most of what it uses will come from the PV on a sunny day.

 

Fit a PV diverter to send excess PV to the immersion heater.

 

That just about covers it for us.

 

BUT if I install any more PV then that will have to have batteries, otherwise it would just be too much to use real time.

Link to comment
Share on other sites

On 06/04/2024 at 14:19, Russell griffiths said:

4.9 kw panels on the roof, how and when will I use any power generated

You can store some energy thermally via resistance heating, that is cheap and cheerful. Did you find out if your ASHP buffer can take an immersion heater? Or plumb in a Willis heater to the UFH.

 

The biggest problem with PV is that on ordinary days, the output is quite variable because of clouds passing overhead. That makes grabbing the energy a bit tricky if you want to reduce your imports. That is why resistance heating is useful, it can react quite quickly.

Running a washing machine is not really a stop/start process, cooking is the same.

It is possible to modify a freezer to start up and over cook when running on PV, but that runs the risk of over cycling the compressor, and these days, freezer use very little power anyway.

 

So I would just divert, via a two stage diverter to DHW and slab heating.

 

For a few quid, it would be possible to fit a relatively small system to run your MVHR as that is not a critical system, and easily overridden manually if needs be.

 

Some battery chemistries are inherently safer than others.

Link to comment
Share on other sites

What kind of FIT can you organise?

 

Even if you can sell at half your buy in rate you might be better off than buying a battery. It just requires some sums. 

Link to comment
Share on other sites

1 hour ago, Iceverge said:

What kind of FIT can you organise?

 

Even if you can sell at half your buy in rate you might be better off than buying a battery. It just requires some sums. 

FIT ended in 2019 and was replaced with SEG that for suppliers over a certain size must pay at a pittance of a rate-5-7p/unit. Both government backed and regulated schemes

 

Octopus offer their own export tariff at 15p/unit, but that can end at anytime as its neither FIT or SEG so no government backing. Loss of that rate is worth considering when calculating battery payback.

  • Like 1
Link to comment
Share on other sites

7 hours ago, NailBiter said:

Consider an energy management such as EMHASS (https://emhass.readthedocs.io/en/latest/intro.html#what-is-energy-management-for-home-assistant-emhass) to perform load shifting of your deferrable loads.

Live looks to short for that. Got a solar forecast app for the battery and that's a pain, might be ok May to August, but the rest of the year a lot to hit or miss to be relied on.

Link to comment
Share on other sites

10 hours ago, JohnMo said:

Live looks to short for that. Got a solar forecast app for the battery and that's a pain, might be ok May to August, but the rest of the year a lot to hit or miss to be relied on.


That would very much depend on personal circumstances including location, insolation, azimuth, pitch, size of the array, size of deferrable loads and your familiarity with software / machine learning.

Link to comment
Share on other sites

Many lithium home batteries are IP65 and can be installed outside. I know the SolarEdge EnergyBank is one of these.

Use load shifting (see above) to keep the battery size as small as you can for your house.

A runaway Lithium fire outside will do damage but depending on what the exterior of your building is made out of will not spread further.

Also don't forget batteries can be installed quite far away from inverters. Ours is all in the shed at the bottom of the garden.

Link to comment
Share on other sites

5 hours ago, NailBiter said:


That would very much depend on personal circumstances including location, insolation, azimuth, pitch, size of the array, size of deferrable loads and your familiarity with software / machine learning.

My point was the solar forecast said I would be doing 18kWh today, so far today with 2hrs of generation to go I have only done 7kWh. Plus to fire up the dishwasher or washing machine someone needs to fill it, machine learning doesn't really help. Possibly 5 months of the year good, rest of the time too hit and miss. 

 

Been caught a few times where I ended up importing expensive electric because machine learning said I could charge the battery to 50% as solar forecast said xkWh  generation expected and it didn't anywhere near that.

 

Nothing to do with location, insolation, azimuth, pitch, size of the array.

 

Used the excess manually today, oven on making granola, washing machine on, dishwasher on. The rest diverted to DHW.

Link to comment
Share on other sites

5 hours ago, JohnMo said:

solar forecast said xkWh  generation expected and it didn't anywhere near that.


Of course it depends on the quality of your forecast data (and how accurately you modelled your array and shading).

It also depends on the optimisation strategy you use, day-ahead would have less ability to respond to forecast error than MPC.

Link to comment
Share on other sites

46 minutes ago, NailBiter said:

Of course it depends on the quality of your forecast data (and how accurately you modelled your array and shading).

Pretty accurate, trouble with winter and spring the variability of the weather and variable usage

 

My forecast (from Solcast) for the next couple of days, the white is the low forecast the pink is max and the orange line the averaged expected. A lot of variability. It also takes zero account of temperature.  Today heat pump barely ran, tomorrow will also barely run, the next day it will, so although the prediction tool looks back (consumption) and ahead (PV output and averaged out consumption) it's a bit of a finger in the air during the heating season.

 

So although home assistant knows what happens in the past and therefore predicts the future likelihoods, the likely scenario on the 9th it would see low consumption on the 7th and 8th build that in to the kWh likely to be consumed on the 9th and give a short fall in SOC at the end of my E7 window. When heating with a heat pump and you get variable weather (we all do), and this part of the picture is missing it isn't or cannot be best placed to accurately predict.

 

@NailBiter what forecasting tool are you using?

 

 

image.png.896107409da794d2f7dd8931fb867c78.png

 

Link to comment
Share on other sites

Does anyone have any links to the statistics on domestic battery fires? AFAIK there's more chance of the tumble drier going up in flames than a 'solar' battery :/ 

 

Put it outdoors in a second 'kiosk', adjacent to the one doing the mains? Metal-clad if possible.

 

Mine will go in the shed / outbuilding, as is my plan for my current clients project; (20kW battery per phase / 3ph + ~16kWp of solar). I've not just looked at risk mitigation for this myself / my client(s) as the kit is sizeable too, so I'd prefer not to lose any of the GIA of the actual dwelling either so will mount / locate remotely. Cables will be oversized from outbuilding > boundary kiosk > dwelling accordingly to minimise voltage drop to the nth degree, as a 35mm2 cable isn't lifechanging sums more than a 25mm2 plus leaves room for 'more' if downstream bulking of battery system is deemed advantageous (~1900m2 residence with indoor swimming pool, sauna / steam-room et-al). I'm putting a guesstimated size system in and will recommend running the home for 12 months to gather statistics, so will always consider a structured battery / solar setup that can grow.

 

I dislike the notion of sizing to take advantage of a particular electricity providers 'offering of the day', seems too risky for me to spend a clients money on (I have responsibilities as the Main Contractor and consultant, so my 'ideas' need to be robust before I convert them into an actual recommendation). I prefer knowns, so work on solar capture and look to store and utilise this pragmatically. IF a favourable provider / tariff then comes along downstream we can add that to the mix to maximise RoI, vs rely on it.

 

@Russell griffiths, defo have the batteries as you and Mrs G work from home, just put them outdoors if you're that worried. Simples.

  • Like 1
Link to comment
Share on other sites

10 hours ago, JohnMo said:

My forecast (from Solcast) for the next couple of days

How did you register for Solcast? If I go on their website, and select Home User, I get switched between 2 web pages and can't register.  Frustrating.....

 

Simon

Link to comment
Share on other sites

11 hours ago, JohnMo said:

Pretty accurate, trouble with winter and spring the variability of the weather and variable usage

 

My forecast (from Solcast) for the next couple of days, the white is the low forecast the pink is max and the orange line the averaged expected. A lot of variability. It also takes zero account of temperature.  Today heat pump barely ran, tomorrow will also barely run, the next day it will, so although the prediction tool looks back (consumption) and ahead (PV output and averaged out consumption) it's a bit of a finger in the air during the heating season.

 

So although home assistant knows what happens in the past and therefore predicts the future likelihoods, the likely scenario on the 9th it would see low consumption on the 7th and 8th build that in to the kWh likely to be consumed on the 9th and give a short fall in SOC at the end of my E7 window. When heating with a heat pump and you get variable weather (we all do), and this part of the picture is missing it isn't or cannot be best placed to accurately predict.

 

@NailBiter what forecasting tool are you using?

 

 

image.png.896107409da794d2f7dd8931fb867c78.png

 


Solcast is reasonably accurate (if you pay for access to the lower interval data), I'm guessing you are using the free high interval data but it is hard to tell from that graph. That is only 1 of the forecasts EMHASS needs to work however. You also need to be accurately forecasting use and that forecast also needs to be accurate.

Your forecasts may be off as your array may be a slightly different azimuth / pitch or your panels may be producing a different amount of power than you told the solar forecasting tool. You might also have shading issues.

At first it is likely EMHASS will be wildly wrong but over time it will learn. Yes that has to be from the past as we can't learn from the future. 

If your claim is that accurately forecasting the above is impossible I'd contend that hard is not the same as impossible.

Using data and machine learning algorithms to calculate something is the exact opposite of finger in the air. Relying on an individual to monitor by manually and respond manually is a much more time consuming and error-prone approach but to each their own.

I think you may have missed that it isn't just using past data to make predictions it uses current data to update those predictions.

You don't have to use it if you don't want to, nobody does, equally it is a tool there waiting to be used if you want to.

Edited by NailBiter
Link to comment
Share on other sites

58 minutes ago, Bramco said:

How did you register for Solcast? If I go on their website, and select Home User, I get switched between 2 web pages and can't register.  Frustrating.....

 

Simon


Try a different browser, I just ran through the registration flow with a disposable email and it worked fine, nothing odd to report.

Link to comment
Share on other sites

23 minutes ago, NailBiter said:

Try a different browser

Nope, still doesn't work - I've emailed support, maybe there's an issue with their systems.

 

Simon

Link to comment
Share on other sites

@NailBiter and anyone else trying to register on Solcast, the site is quite confusing.  The registration page has 3 boxes with text in them for each type user user. Inside each box, there is a hyperlink but this link takes you to an information page about the service for that type of user. So you end up in  a loop.  You need to click anywhere in the box for the different user types - it's not obvious that the box itself is a hyperlink.

 

I've suggested to them that this is a little confusing.

 

Simon

 

PS @NailBiter thanks for your offer to help by DM

  • Like 1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...