Ed Hart
Member
Wow....that is very complicated. I think my contractor made sure that I would not be confused: Set your minimum charge level to 20%, and only charge with solar.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Time of Use Power Shifting for Powerwall 2
- Thread starter zanary
- Start date
-
- Tags
- Powerwall Tesla Energy
Not a great idea - you will never charge from solar - in a region where there is no NEM, you will end up exporting a lot of energy for very little value.
In Australia there is not ITC so Powerwall can already charge from the grid. It sounds like you are asking for Time-Based Control - you should call Tesla and ask for it to be enabled (if it is not currently available).
Are you sure it is not maximising the value of your solar? IE if you have a lot of excess solar - you DO want to charge from that (free), rather than the grid (cheap)...
In my mind free > cheap
If you get nothing for exported solar, true. In my case (and I suspect @poindexter's) I get paid 11.3cents for exported energy, but pay 9.8cents for off-peak grid - so it is definitely better to charge from the grid overnight to full, then allow the PW to run in self consumption mode to offset peak consumption, recharge from PV, then export.
@poindexter wants to take this one step further - charge overnight, offset consumption during the day (but not recharge) so all solar is exported, and charge again at night.
In my case, it's only worth 13.5 x 1.5 cents a day (diff between FIT and OP), but still, $73 per year will buy me a beer a month. Better in my hands.
Final thoughts - in VIC from next July we'll have FIT = 29 cents, OP = say 9 cents. Then, this could be worth $985 per annum.
I think I'll have to test this theory...
Don't forget round trip efficiency
- every kwh you charge from the grid, you'll only get 0.9 back - so your off-peak rate effectively becomes 10.8cents - so you are saving 0.5cents per kwh or $24 per year - 2 six-packs is down to 1 !!
Also don't forget that that required a full 13.5 kWh of peak usage to discharge into, which is actually quite a lot for just a few hours in the eveneing
Also don't forget that weekends don't have the same spread as weekdays... seem like a lot of work for little upside - might get 1 beer out of it.
With a much higher FiT - make a lot more sense... But generally self-consumption of solar, with priority for any remaining grid consumption to occur during off-peak it probably the best way to go.
True, but similarly, every kWh you charge from solar, you only get 0.9 back as well - so that 1 kWh of storage has cost me 12.4c worth of feed in - so I'm back to more than one six-pack. And, trust me, with a full household, an electric oven, dishwasher, A/C in summer, heater blower in winter, and peak that runs all the way through to 11pm, it's not hard at all to chew through most of 13.5kWh after sunset.
We are still getting to know our PW2 + PV (in Melb, Australia, peak 7am to 11pm weekdays, other off peak) and the bit missing seems to be predictive charging - so dependent on predicted cloud cover what charge should it be carrying at 7am? On sunny day ahead should be low at 7am, but on a cloudy day should be high. In TOU - Cost Saving it seems to manage to backup reserve level at 7am (10% in our case) independent of forecast. Other than that use of solar storage / export/ use appears sensible (see attached for sunny day) -- seems to start topping up around 11pm with a view to getting to low point around 7am.
Cloud / solar forecast data is available. Best alternative at present seems to be to put it in backup only mode around 11pm with a set reserve level (roughly aligned to cloud forecast) and put it back in TOU in the morning. I am guessing others have found a way to automate this.
Cloud / solar forecast data is available. Best alternative at present seems to be to put it in backup only mode around 11pm with a set reserve level (roughly aligned to cloud forecast) and put it back in TOU in the morning. I am guessing others have found a way to automate this.
For the whole arbitrage charge at night and discharge at peak game --- that's cute, but we took the 30% ITC money instead since SGIP still requires a high solar rate anyways, wasn't worth trying to game the system.
We live in SDGE country and just found out about TOU-5 rate schedule mentioned above. With our undersized solar since 2014 and now EVs and the powerwall since April, I switched over last night to TOU-5 despite being NEM 1.0 and Grandfathered to TOU-2 until 8/2019.
Simply put, the $0.09 cheap super-offpeak rates overrides everything. The PV will cover 4p-6p peak times and probably feed a little high NEM back to the grid and energy shifting that Powerwall allows will cover the 6p-9p usage easily for us.
I'm just glad they made cheap super-off-peak rates again available like it was in the initial EV pilots. Energy is dirt cheap at night since they have to maintain the grid 24/7.
We live in SDGE country and just found out about TOU-5 rate schedule mentioned above. With our undersized solar since 2014 and now EVs and the powerwall since April, I switched over last night to TOU-5 despite being NEM 1.0 and Grandfathered to TOU-2 until 8/2019.
Simply put, the $0.09 cheap super-offpeak rates overrides everything. The PV will cover 4p-6p peak times and probably feed a little high NEM back to the grid and energy shifting that Powerwall allows will cover the 6p-9p usage easily for us.
I'm just glad they made cheap super-off-peak rates again available like it was in the initial EV pilots. Energy is dirt cheap at night since they have to maintain the grid 24/7.
Yes. My approach (since this is what I knew - I've long since stopped being a developer) was to write a windows service that gets the solar forecast for the next day, anticipated load data, and sets reserve percentages to get it to charge as much as required when required. Setup isn't as dead simple - there's no gui to set it up, but it does the job nicely. Because my FIT is > that my OP, I prefer standby over discharging into OP (more to export to FIT), and prefer over-charging to under-charging (over charging and exporting some FIT is better than under-charging and using some peak). See Flat Out 4.840kW | Live Output (extended parameters view) to see how it operates.
P.S. you need to go back to Flat Out 4.840kW | Live Output or earlier to see it charging - it hasn't needed to for the last week with the weather we've had.
That looks to do be doing exactly the right thing. Nice work. I have only just started looking through the API - but looks like you have done all the hard work!
For the whole arbitrage charge at night and discharge at peak game --- that's cute, but we took the 30% ITC money instead since SGIP still requires a high solar rate anyways, wasn't worth trying to game the system.
We live in SDGE country and just found out about TOU-5 rate schedule mentioned above. With our undersized solar since 2014 and now EVs and the powerwall since April, I switched over last night to TOU-5 despite being NEM 1.0 and Grandfathered to TOU-2 until 8/2019.
Simply put, the $0.09 cheap super-offpeak rates overrides everything. The PV will cover 4p-6p peak times and probably feed a little high NEM back to the grid and energy shifting that Powerwall allows will cover the 6p-9p usage easily for us.
I'm just glad they made cheap super-off-peak rates again available like it was in the initial EV pilots. Energy is dirt cheap at night since they have to maintain the grid 24/7.
You can take the 30% ITC and still apply for the SGIP rebate as well (have your cake and eat it too). The only stipulation on SGIP is you are agreeing to discharge/energy shift a certain amount per year (for 5 years) instead of using it in backup-only mode (and the discharge doesn't have to be during peak). Since you are using it for energy shifting anyway, you might as well apply or you are just leaving money on the table.
Yes, already applied, Step 4. had waited on pulling the trigger on the powerwall in hopes of V2H solutions and missed the first three steps; finally gave up and took the dive on the powerwall at Step 4. Probably won't get the $$$ back until 2019 since only Step 2 folk getting theirs at this time.
It's par 0.90 where it's PV or grid charged so you can eliminate that argument.
The math after that is whether PV and losing FiT is better to charge the battery with than grid, over the entire 10 year lifetime.
I tried doing 88 simulations by hand in NREL's System Advisor Model with different battery permutations of variables: lowest SoC allowed (0 to 80% in increments of 10%); charge before load; feed load before charging; charge via grid only on off-peak; charge via PV during TOU periods of 1, 1+2, 1+2+3, or 1+2+3+4, 2, 3, 4 (1=off-peak, 2=shoulder, 3=peak, 4=after-peak-shoulder). Rates are SCE's TOU-A.
Generally, charging off-peak via grid saves the most money, fastest payback on system-costs, and gives the highest net-present value just due to the lower base-cost.
Charging via PV and limiting discharge period to only the most expense TOU periods is next best but depth of discharge must be limited or it costs more to PV-recharge the battery next day.
For example, unlimited PV-charging before house-load, but discharging only in period 3, the best SoC is drain down to 10%, but saving a "$5053" out of $5050 annually. This can be adjusted to lowest SoC of 50% and still saving $5051 out of $5050.
Whereas, if feeding house-load before charging battery, it's again 10% as best lowest SoC, but saving a "$5094" out of $5050 annually. However, lowest SoC can be 70% and still saving "$5067" out of $5050.
Comparatively for grid-charging, discharging only at peak to max low of 10% saves "$5129" out of $5050 annually.
Free-discharging at periods 2,3,4 to max low of 10% SoC saves "$5834" out of $5050. Alternatively, to 80% SoC saves still "$5049" out of $5050 -- so grid-charging seems the best way to go to lower-the bill, while also preserving battery-life.
Can people chime in and indicate whether or not the Time-based Control is available and where they are located?
I am in Colorado and recently turned on my Powerwalls. Originally, I was on 1.17.x before I turned them on but got upgraded to 1.22.1 recently. I do not have Time-based Control. From what I've read, it looks like it is only available in California and Australia.
Any idea on when it will be available in Colorado (or elsewhere)?
I am in Colorado and recently turned on my Powerwalls. Originally, I was on 1.17.x before I turned them on but got upgraded to 1.22.1 recently. I do not have Time-based Control. From what I've read, it looks like it is only available in California and Australia.
Any idea on when it will be available in Colorado (or elsewhere)?
Confirm available in Victoria, Australia - but I needed to ask support for it to be enabled.
Shygar
Member
If you login with your customer credentials to your web based access for the powerwall/gateway, you will see the mode you are set to. It should say TOU something on that mode.
