Time-Based Control?

[Ampster]

I don't discharge after 7pm because I use OhmConnect during that time)

w
When I was load shifting with a hybrid inverter and had OhmConnect I rarely got any money back because my baseline was already low during Ohmhours. I never did the calculation to see what would happen if I gamed the system, but I have heard of situations where some people are getting lots of incentives.

 

[cwied]

I never did the calculation to see what would happen if I gamed the system, but I have heard of situations where some people are getting lots of incentives.

I think calling it "gaming the system" is overly negative. Reducing grid usage when energy is plentiful is much less useful than reducing when demand is high. In practice, I wouldn't usually have enough energy to discharge during the evening part-peak period anyway.

Longer-term I hope Tesla does implement a real grid-stabilization program that I can participate in. Until then, my participation in OhmConnect on the whole has a positive impact on the grid since I'm definitely reducing my load during the critical hours.

 

[arnolddeleon]

...
Longer-term I hope Tesla does implement a real grid-stabilization program that I can participate in. Until then, my participation in OhmConnect on the whole has a positive impact on the grid since I'm definitely reducing my load during the critical hours.

I'm hoping too, I think this is where the real money will be.

 

[arnolddeleon]

...
I'm going to let the system blunder through for now so the support team has something to look at tomorrow.

I little gauche, quoting myself. But anyhow the system did discharge from 5 to 8 pm (part peak).

 

[arnolddeleon]

Tier 2 support won't look at the case unless tier has had you power off the powerwalls (switch and breaker) for 3 minutes.

 

[Artimis]

w
When I was load shifting with a hybrid inverter and had OhmConnect I rarely got any money back because my baseline was already low during Ohmhours. I never did the calculation to see what would happen if I gamed the system, but I have heard of situations where some people are getting lots of incentives.

I've been having a really hard time with OhmConnect also ever since I installed solar + 2 Powerwalls. I've heard it's easier for people to "game the system" if they have an EV that they routinely charge during a certain time of day that coincides with an OhmHour. They simply don't charge at that time.

Currently, I am still grandfathered into tiered pricing and normally operate on Self-Consumption mode. I have my OhmHours setup from 5pm - 7pm. Normally, my wife is cooking around those times and we have a lot of appliances going and since the sun is low, our solar excess is next to nothing allowing for a low baseline (e.g. 0 to -.2 kWh).

When an OhmHour is coming, I switch to "Cost-Savings" mode with peak hours setup from 4:30-8pm (no shoulder defined). This forces the gateway to run the house entirely off the Powerwalls and export ALL solar production into the grid (e.g. 1.6 - 0.6 kWh). Once the OhmHour is over, I switch back to Self-consumption mode.

This plan can still backfire on a cloudy day and hours will need to be adjusted during winter but it seems to be working so far.

 

[cwied]

@Artimis - the key to OhmConnect is that it's a targeted demand-reduction program. If you already reduce your demand on all days by using self-consumption mode, then you won't have any demand to reduce on the target days. Therefore, if you set your OhmHours to 5pm-7pm, you should have the Powerwalls on standby (or charging) during those hours if you want to participate in OhmConnect. If you don't want the risk of cloudy days, you can set your OhmHours later in the day to avoid solar production fluctuations. I have my OhmHours set from 7pm-9pm and that's why I've set those hours to off-peak in my schedule.

Out of curiosity, what is your motivation to run on self-consumption mode? Given the round-trip efficiency losses, don't you end up paying for more electricity than if you were to just use net metering?

 

[arnolddeleon]

Last Friday my Powerwalls did not discharge at all is Cost Saving mode. Today is Friday and they are working "correctly". Well at least consistent for the week. They're not quite perfect yet since they start discharging during the part peak in the morning and the system is running out of available capacity before the peak ends. So it would make more sense to not discharge until peak time. They are also not charging during the part peak even though there is still room in the battery.

 

[cwied]

I think the not charging to 100% is intentional. My theory is that they allocate a portion of the battery to peak energy and a portion to part-peak energy. If the battery is over the peak threshold (80%, for example), it'll discharge to that level in the part-peak period. If it's under, it'll charge up to that level on part peak. It'll always charge during the off-peak period and always discharge during the part-peak period.

The end result is on the weekend when there's plenty of off-peak power, it'll charge to 100% and then discharge during the part-peak period. During the weak, it'll charge to the intermediate threshold during the part-peak period and then discharge during peak.

Note that last week the threshold it seemed to be charging to was 80%. This week it seems to be 85%, so it probably adjusts to usage.

There is some behavior that is not explained by my current model of their algorithm. For instance, for some reason the batteries started discharging a little (maybe about 30 minutes) before the peak started this week. It's probably more complicated than what I described.

Also, the Tesla support person insists that she doesn't expect my Powerwalls to charge during the weekend at all, because "they won't charge unless there's a peak on that day." I do currently have my weekend periods defined with a fake peak for the last half hour to see if Cost Saving is more stable that way.

 

[arnolddeleon]

I think the not charging to 100% is intentional. My theory is that they allocate a portion of the battery to peak energy and a portion to part-peak energy. If the battery is over the peak threshold (80%, for example), it'll discharge to that level in the part-peak period. If it's under, it'll charge up to that level on part peak. It'll always charge during the off-peak period and always discharge during the part-peak period.

The end result is on the weekend when there's plenty of off-peak power, it'll charge to 100% and then discharge during the part-peak period. During the weak, it'll charge to the intermediate threshold during the part-peak period and then discharge during peak.

Note that last week the threshold it seemed to be charging to was 80%. This week it seems to be 85%, so it probably adjusts to usage.

There is some behavior that is not explained by my current model of their algorithm. For instance, for some reason the batteries started discharging a little (maybe about 30 minutes) before the peak started this week. It's probably more complicated than what I described.

Also, the Tesla support person insists that she doesn't expect my Powerwalls to charge during the weekend at all, because "they won't charge unless there's a peak on that day." I do currently have my weekend periods defined with a fake peak for the last half hour to see if Cost Saving is more stable that way.

Hmm, I like your theory. I think early on I saw it discharge and charge during part peak.

I think they are trying to predict usage. I don't know what is in their model (e.g. are they factoring in weather?). Today even though the system didn't charge to 100% it didn't get to the reserve threshold before the beginning of off-peak. That's because today the weather was not as hot so no air conditioning load.

I don't have a peak in my weekend schedule, we'll see if they charge tomorrow.

 

[arnolddeleon]

Well it looks good so far today, cost savings charged today and I'm now in part-peak and it is discharging. I haven't received an explanation as to why it appears to be working mostly correctly now.

 

[ChrisMPK]

I had the same experience last Friday (6/15). I'm on 1.15.8 with TBC set to cost saving. It has been working as expected, charging during the morning shoulder, then discharging from peak through the evening shoulder. This leaves the PW2 at about 60% and it gets back to 100% in the morning.

Coincidentally, I had changed the reserve from 15% to 25% some time on Thursday as I was never getting down that low. I'm not sure if disturbing the reserve triggered the Friday hiatus or if I'm seeing the same behavior as arnolddeleon. It also didn't discharge at all on Saturday and today (Sunday) it has been discharging during the morning off peak and just stopped when part peak arrived (no peak on weekends in SCE land).

 

[arnolddeleon]

I had the same experience last Friday (6/15). I'm on 1.15.8 with TBC set to cost saving. It has been working as expected, charging during the morning shoulder, then discharging from peak through the evening shoulder. This leaves the PW2 at about 60% and it gets back to 100% in the morning.

Coincidentally, I had changed the reserve from 15% to 25% some time on Thursday as I was never getting down that low. I'm not sure if disturbing the reserve triggered the Friday hiatus or if I'm seeing the same behavior as arnolddeleon. It also didn't discharge at all on Saturday and today (Sunday) it has been discharging during the morning off peak and just stopped when part peak arrived (no peak on weekends in SCE land).

How long have you been on TBC? When I experienced the no discharge Friday it was the first week for me. I don't know how long @cwied had TBC when he experienced it. My Powerwalls are currently at version 1.16.2.

This weekend the system has been behaving as expected so far. It discharged only during the part-peak yesterday and it charged to 100% this morning.

 

[ChrisMPK]

I've only had it a week or so. Today it was discharging until part peak at 8am and hasn't recharged yet.

 

[cwied]

I had already had TBC for several weeks when I had the issues, and I had them for more than one week consecutively. I do switch modes relatively frequently, though, so I don't know if that caused any of the issues. It does seem to be an issue that is happening to many people, though.

My system has been behaving as I would like it to on cost saving for the past week, though, even with the mode switching. I hope it stays that way.

 

[lance.cazeault]

I had already had TBC for several weeks when I had the issues, and I had them for more than one week consecutively. I do switch modes relatively frequently, though, so I don't know if that caused any of the issues. It does seem to be an issue that is happening to many people, though.

My system has been behaving as I would like it to on cost saving for the past week, though, even with the mode switching. I hope it stays that way.

New member to the forum, so I'm still reading through and catching up on previous posts.

I saw in your (cwied's) posts that you are experiencing issues similar to what I am.

I've had the TBC option available to me for i think about 7 weeks now (possibly a little longer). Starting in late May it started acted screwy.

The last week of May or so, the PW quit charging to 100% (after calling a couple of times, I was finally told by a TESLA rep that Tier 2 had examined my PW, and everything looks good. The PW was operating as designed, and it not fully charging is operating the way it needs to for the Cost Saving setting. (The explanation seemed odd - i would think that the PW would want to charge fully because clouds and other weather factors could always change). I hit the "BELIEVE BUTTON" and moved on continuing to watch it.

Starting Friday (JUN 1), my PW stopped charging at 0800. It resumed charging between 1330-1530 (reaching 100%). However it did not discharged any during my PEAK period (set for 1530-2000) --- this behavior also occurred on Friday, JUN 8 and Friday JUN 15.

On Saturday (JUN 9), the PW did not charge during the OFF-PEAK period. It actually started charging during the Shoulder period and did not discharge at all. This behavior also occurred on Saturday, JUN 16.

On Sunday (JUN 10), the PW began discharging at Midnight and continued to discharge until it reached the reserve limit that I set. It did not charge at all on Sunday.

On Mondays, it begins charging and discharging per schedule and everything works perfectly until the next Friday.

I called on Saturday (15 JUN) and was told that this issue is at Tier 2 level and is being worked.
(Even though I know it has cost me a little in savings, I have not changed any setting on my system since 27 May. I wanted the TESLA techs to see things as they are without the system trying to change things).

 

[cwied]

See my post further above as to why I think not charging fully on part-peak energy makes sense. I can't think of any good reason not to charge fully off-peak, though.

I'd be interested if your Powerwalls self-correct as arnolddeleon's did. I'm also interested to see if mine behave correctly two weeks in a row.

 

[lance.cazeault]

Monday rolls around, and the Powerwall works as it is expected to.

I'll let you know what happens come Friday.

 

[arnolddeleon]

I've had about 10+ days of pretty solid TBC behavior. The system seems to be learning, the last two days it didn't start discharging until the peak (it was starting about 30 minutes too early before). Today it did a little charging during part peak as well.

I think my system is going to turn out be a stress test for cost control mode. On "nice" days (where we don't run AC) I have more than enough battery to go through the peak. On those days it would have made sense to start discharging during first part peak to absolutely maximize savings. There wouldn't be enough load on the second part peak to use up the battery. OTOH on warmer days, the system needs to reserve all the battery capacity for peak. The only way the PW can really get this right to to have weather as input to their prediction model.

 

[lance.cazeault]

Friday is here ... and it looks like it is going to be another weekend of abnormal behavior from my Powerwall.

If it continues as I expect it to, I'll be back on the phone with tech support again today to continue being that squeaky wheel. (When I called a couple of days ago, I was told I was told that Tier 2 had elevated the ticket up to Service Engineering and had to request a status update from them.) Will find out if there is any update from SE.

Of course with this being the 4th Friday in a row, this is starting to turn into the normal.