I am now on amber - going well and so far much cheaper (19c prior 15c now average)
I have updated all my automations to not run or turn off if the price spikes above 40c (i just set a virtual switch called Price Spike) - (which is about 30 minutes a day some days not often) need more time to see how it works.
Here's what i'm doing:
Lighting automation - do not turn off but allow turn off only if price spiking switch on
Cooling/heating - does not auto turn on if price spike on
fans/other automations - disable if price spike
What i am thinking of adding:
If amber alert me of an upcming spike like $3.00 or more or even $17
Put house into away mode essentially.
How I have it now - users can override but if price is that high i'd like to auto shut off anything that is turned on i.e tv/cooling/lights - pretty easy but wife may not like it
if we get forecast pricing - this will only be really valuable when the api is more reliable i fear... but it would be good to still start cooling the house earlier if a price of 40 cents of so is expectedin next 90 mins.
Issues:
the amber api is delayed by 20-30 mins sometimes. Last night the price went to 43 c per hour and the api was showing 20 for a good 20 minutes before it caught up. (I have asked amber if they are improving current api at all or do we need to wait for full public... guessing the answer is the latter
Apparently there is a 5min price and 30 min price. Hence the delay. Nice.
Going to try this out here's the detail from amber:
During a 30 minute period, the price is updated every 5 minutes in the energy market. These are returned as 5MIN prices in the API. During a 30 minute period, the live price shown in the app is an average of these 5MIN periods (we collect each of the 5MIN prices and average them). At 25 minutes into a 30 minute period , the price is finalised and is returned as a 30MIN price in the API.
So at 12:20, the latest 30MIN price will give you the price at12:00 (you should actually be using the 5MIN price). At 12:25, the latest 30MIN price will give you the 12:30 price (you should be using the 30MIN price).
In the code you shared, you're only using the 30MIN price, never the 5MIN price. This explains why your data is always at least 25 minutes out of date.
If you update your code to use recent5minPeriod when available, you should get a result closer to the app. It won't be perfect as this won't perform the average of all 5MIN prices like the app does.
Ok, good coding
Do you still want me to look at the averaging part - looking at how many minutes into the 30 minute period it is, or shall we just leave it with the change you made? I can then concentrate on doing the forecast prices as extra attributes like we discussed.
I have solar, but only about 2.4kW. The feed-in is often quite low, around 4c, and really only goes up above 10 at the end of the day when I'm only generating 400W or so. But in Victoria we get the best of both worlds as Amber will make sure we always get at least 10.2c. They calculate how much we lost out and credit it back each July. Nice!
it's weird how they have done that. i don't really want to have to check two different values. i just want a current price you know? then run all the rules off that.
Sounds like my first system back in Adelaide - for our current house I went as big as I could afford (5.4kW) with the best components I could afford at the same time (Fronius Inverter and LG panels).
That is cool! I'm definitely going to do some data modeling to see if it's worth me switching!
2.4 is better than none. i am waiting for our extension to be done (adding another level) so will put either tesla solar tiles if they are out finally (we will be building in 3 years) or just solar. cannot wait hopefully tesla power walls are a lot cheaper by then as well.
I'd love to re-do my roof in these at some point - they seem to be a ways off tho.
You definitely need 5+ kW of panels (preferably more) to make them worthwhile. I calculated that I'd save another $700 per year If I had one - unfortunately that just doesn't add up economically at $12k install.
My Brother-in-law in Adelaide got his for under $5k in a deal with AGL to be part of their VPP - I wish there was a deal like this here In Victoria.
Well I'm excited I just had a spike of $0.30 and that's my threshold so everything shut down. And then went back to normal after the spike ended . Lovely!!
The reason we only got 8 panels up there earlier this year was due to the shape of our townhouse roof and its east-west orientation. We could also get another 9 facing south which isn’t actually as bad as it sounds, particularly in cloudy Melbourne. We got Enphase microinverters which limit the headline figures (295W per 350W panel) but work brilliantly in practice.
I'm seriously considering adding another 1.5 kW array to our SW (220°) roof to extend our generation late in the day. Currently, we have 1.5 kW array facing NE (45°) which works great in the morning (typically meeting our energy needs by 8 am on a sunny day), and 3.9 kW array facing NW (315°).
Our Inverter is only 5 kW so I'd get some mid-day clipping which I can live with for the improved overall generation numbers and late day coverage. Here in the Yarra ranges it tends to be hotter in the late arvo, rather than mid-day (like it is in Adelaide).
Thanks so much for this, working somewhat like a charm! I've noticed a 1c discrepancy so far (hubitat is saying it's 14c when amber is saying 15c), I'm guessing this might be a decimal related thing?
Yes, it never quite matches what the app shows, but the app is using a different data source that we don’t have access to yet. I’m hoping we’ll be able to use that later this year some time. That’s when I heard they will make a proper supported API.
