Tesla Virtual Power Plant in CA

[sborsch]

Just joined tonight since I was approved and am now active in tonight’s VPP event (my second one) but am uncertain on something in which someone may have some insight.

In the first image below, I see all the CAISO data of demand and supply in their app. But the second image, which is my Tesla app showing the VPP event and the SCE Fleet Homes like mine, show a reserve just under 7MWs.

So my question is this: Does SCE and PG&E Fleet Homes’ stored energy appear in CAISO’s data? Or because the VPP is still in ‘beta’ in California, it’s not yet reflected in any way?

 

[cwied]

Just joined tonight since I was approved and am now active in tonight’s VPP event (my second one) but am uncertain on something in which someone may have some insight.

In the first image below, I see all the CAISO data of demand and supply in their app. But the second image, which is my Tesla app showing the VPP event and the SCE Fleet Homes like mine, show a reserve just under 7MWs.

So my question is this: Does SCE and PG&E Fleet Homes’ stored energy appear in CAISO’s data? Or because the VPP is still in ‘beta’ in California, it’s not yet reflected in any way?

View attachment 849748

View attachment 849749

The number at the bottom of the Tesla screen is how much power the fleet is producing. That's 6.8 MW so it isn't really big enough to be noticeable in the CA ISO data.

 

[Redhill_qik]

Here is the PG&E VPP Fleet Home data that I collected for the 9/6 event at 10 minute intervals. At 4:00pm when charging stopped it was at 742 kW, so not all Powerwalls were at 100%. The peak discharge was at 6:10pm at 23,801 kW with an estimated 46.3 MWh delivered during the event with 3,514 homes.

There were more homes in this event than in the prior even (3514 vs 3462) t, but the peak discharge was lower (23,801 vs 24,711). A couple of thoughts on why this might be:

• Might have pushed reserve levels to 100% to not opt-out, but also to not discharge
• Homes that opt-out for the event don't reduce the fleet homes count
• Some homes may have been off-grid due to local outage
• Home decided to conserve more which reduced the amount that the Powerwall could discharge due to site export limit

I was in the last camp, I kept our A/C off during the event partly to help out and mostly because I was still enrolled in SmartAC and PG&E was disabling my A/C. This reduced my Powerwall discharge from ~9.8kW to ~8.0kW so that I would not exceed my PTO limit of 7.86kW. I had raised my reserve from 5% to 30% due to nearby outages during the event, but then drop it again in the final 15 minutes and ended up at 24% with 19.7 kWh discharged vs 24.7 kWh on 9/5.

 

[mjhwa]

Hey guys maybe this is a dumb question, but I just joined VPP and went through a couple of events, but how do I tell how much I’m being paid after each event ends?

If you like to tinker, you could use the Tesla API to get this value:

grid_energy_exported_from_battery

It can be retrieved for All Day, Off Peak, Partial Peak, and Peak depending on what mode your system is in. For PG&E the system is blocked from exporting to the grid from your Powerwall, except for VPP events. So if you retrieve that data you'll notice that most days the value is 100W or less but during the VPP events it's above 1,000W. So you could estimate your payment by taking the value, divide by 1,000W, and multiply by $2/kW.

 

[aesculus]

Here is the PG&E VPP Fleet Home data that I collected for the 9/6 event at 10 minute intervals. At 4:00pm when charging stopped it was at 742 kW, so not all Powerwalls were at 100%.

I am still a bit confused on the prep for a VPP. So the lower graph is probably PWs being charged by excess solar? If so that would imply that by 4PM the batteries would have all reached 100%?

On Mondays event I was at 70% at 1:30 (peak starts at 2PM - well that was not really true since it was a holiday but the Tesla system thought so). The algo in the app decided I needed to have my PWs charged to 100% before peak started and solar was not going to get that done so it grid charged the PWs for that 1/2 hour to get me to 100% at peak. You can see this in the finger of the graph.

Note my return rate is capped at 10 kW and during a VPP, it just dumps it all at max rate until it hits my 20% reserve.

 

[Redhill_qik]

I am still a bit confused on the prep for a VPP. So the lower graph is probably PWs being charged by excess solar? If so that would imply that by 4PM the batteries would have all reached 100%?

On Mondays event I was at 70% at 1:30 (peak starts at 2PM - well that was not really true since it was a holiday but the Tesla system thought so). The algo in the app decided I needed to have my PWs charged to 100% before peak started and solar was not going to get that done so it grid charged the PWs for that 1/2 hour to get me to 100% at peak. You can see this in the finger of the graph.

Note my return rate is capped at 10 kW and during a VPP, it just dumps it all at max rate until it hits my 20% reserve.

Most of your Powerwalls storage was recharged from solar (the yellow part), but some of the recharge happened from the grid (the grey part below the yellow part). So you must have "Grid Charge" enabled and since the Tesla App thought it was a regular weekday vs a holiday and Peak started at 2:00pm it decided to pull from the grid to get you to 100% at the lower rate so that your solar could be sold back during peak at the higher rate.

If your PTO export limit is 10kW or higher and there is no solar production then yes two Powerwalls would discharge at the max 5.0kW rate. You should have still had some solar production at 6:00pm, so I think that your PTO export limit is more than 10kW.

 

[tivoboy]

While the deployment of Powerwalls is currently very low, relative to overall Grid needs (in a state like CA at least for sure), I see a day in the not too distant future (say 10-12 years) where EVERY car is connected to a home and drawing power from a renewable powered grid during peak power generation (06:00-18:00) periods, taking battery in the vehicle up to 80-100% depending on if the vehicle is participating ..then, the grid will simply pull that power back out of the VEHICLES (8-10x the storage capacity of most Powerwalls) in the overnight or peaked periods when needed, overall reducing the need to switch to either gas or coal fired peaked generation plants or providing a significant amount of the needed peak or overnight power needs.

Has anyone ever done a calculation of what say 15M ~85KW capacity vehicles could provide for an overnight or peak demand event might require?

I could see at some point CA requiring all vehicles sold in the state to be able to power back to the home or provide power into a. VPP type of intelligent grid scenario in order to be eligible for CA BEV tax credits,

At that point anyone could provide kWH back to the grid and get some form of compensation.

 

[Redhill_qik]

If you like to tinker, you could use the Tesla API to get this value:

grid_energy_exported_from_battery

It can be retrieved for All Day, Off Peak, Partial Peak, and Peak depending on what mode your system is in. For PG&E the system is blocked from exporting to the grid from your Powerwall, except for VPP events. So if you retrieve that data you'll notice that most days the value is 100W or less but during the VPP events it's above 1,000W. So you could estimate your payment by taking the value, divide by 1,000W, and multiply by $2/kW.

That isn't how the VPP compensation works as it isn't how much is exported to the grid during the event hours from the Powerwall. The compensation is on the difference between the Powerwall discharge during the event hours and the baseline average for the same hours in the preceding days (10 weekdays or 4 weekend/holiday days).

 

[sunwarriors]

While the deployment of Powerwalls is currently very low, relative to overall Grid needs (in a state like CA at least for sure), I see a day in the not too distant future (say 10-12 years) where EVERY car is connected to a home and drawing power from a renewable powered grid during peak power generation (06:00-18:00) periods, taking battery in the vehicle up to 80-100% depending on if the vehicle is participating ..then, the grid will simply pull that power back out of the VEHICLES (8-10x the storage capacity of most Powerwalls) in the overnight or peaked periods when needed, overall reducing the need to switch to either gas or coal fired peaked generation plants or providing a significant amount of the needed peak or overnight power needs.

Has anyone ever done a calculation of what say 15M ~85KW capacity vehicles could provide for an overnight or peak demand event might require?

I could see at some point CA requiring all vehicles sold in the state to be able to power back to the home or provide power into a. VPP type of intelligent grid scenario in order to be eligible for CA BEV tax credits,

At that point anyone could provide kWH back to the grid and get some form of compensation.

A big question is will car makers warranty batteries that have been discharged/recharged all the time for these VPP events? I think there is always some battery degradation (it's no infinite), but cycling the batteries more as a daily thing will seem to increase warranty claims for car batteries, but no benefits for car makers at all.

It's one thing when there is a power outage and the Ford Lightning powers your home for a few days/a week, but another when a consumer is collecting a VPP check, and putting more use out of a battery that has an 8 year warranty. It's still early days so too early to see how this plays out.

 

[tivoboy]

A big question is will car makers warranty batteries that have been discharged/recharged all the time for these VPP events? I think there is always some battery degradation (it's no infinite), but cycling the batteries more as a daily thing will seem to increase warranty claims for car batteries, but no benefits for car makers at all.

It's one thing when there is a power outage and the Ford Lightning powers your home for a few days/a week, but another when a consumer is collecting a VPP check, and putting more use out of a battery that has an 8 year warranty. It's still early days so too early to see how this plays out.

My thinking would be that with a SMART GRID there should be some SMART utilization of batteries or vehicles connected TO the grid. so no, it’s not every VPP event, and not every day and one is not included in every peaker or demand based event. Technically with a SMART grid, the system should cycle through various participants over time so that no one or many participants are participating frequently. Technically a LOW USE EV SHOULD have some type of battery fill/drain cycle going on, and this should be a priority. then next, possibly mobile “power plants” since moving from Home to WORK (charge at work) and drain from say 4-6, then LEAVE work and come home and recharge would be optimal use and care and feeding of the battery.

But, as I said above, some compensation for this type of participant should equalize some of the overall possible degradation that might arise say over a 7-10 year period?

But no, not all the time not every event not draining from full to 10% then to full again - that would not be in the cards. But, I expect SOME capability to be developed and enabled in the next 10 years to optimize this for all parties.

 

[jjrandorin]

My thinking would be that with a SMART GRID there should be some SMART utilization of batteries or vehicles connected TO the grid. so no, it’s not every VPP event, and not every day and one is not included in every peaker or demand based event. Technically with a SMART grid, the system should cycle through various participants over time so that no one or many participants are participating frequently. Technically a LOW USE EV SHOULD have some type of battery fill/drain cycle going on, and this should be a priority. then next, possibly mobile “power plants” since moving from Home to WORK (charge at work) and drain from say 4-6, then LEAVE work and come home and recharge would be optimal use and care and feeding of the battery.

But, as I said above, some compensation for this type of participant should equalize some of the overall possible degradation that might arise say over a 7-10 year period?

But no, not all the time not every event not draining from full to 10% then to full again - that would not be in the cards. But, I expect SOME capability to be developed and enabled in the next 10 years to optimize this for all parties.

This thread is not about theoretical V2H or V2G applications, but for people participating in the current virtual power plant to discuss settings, setup etc. If you want to have a v2h or v2g discussion, I will direct you to one of the existing threads on that topic like this one:

A2H along with Powerwalls

Hi all, Apologies if this has been answered (I searched archives but did not find the exact same scenario) We have 2 PWs with solar running on the grid. I'd like to eventually add an EV with bidirectional charging to this setup (ie, Ford F150 down the road or CT in theory). Would I be able...

teslamotorsclub.com

or this one:

V2G support in Backup Gateway/Switch

According to a reddit post (), the standard for V2G has finally been published. Most people don't realize that a bunch of hardware will need to be installed house side to use V2G. But we do! In theory, Tesla could add support for V2G protocols to the Backup Gateway and PW+ (since that's the...

teslamotorsclub.com

(by "I will direct" I mean that any further posts on that topic will be moved into one of those threads).

 

[tivoboy]

This thread is not about theoretical V2H or V2G applications, but for people participating in the current virtual power plant to discuss settings, setup etc. If you want to have a v2h or v2g discussion, I will direct you to one of the existing threads on that topic like this one:

A2H along with Powerwalls

Hi all, Apologies if this has been answered (I searched archives but did not find the exact same scenario) We have 2 PWs with solar running on the grid. I'd like to eventually add an EV with bidirectional charging to this setup (ie, Ford F150 down the road or CT in theory). Would I be able...

teslamotorsclub.com

or this one:

V2G support in Backup Gateway/Switch

According to a reddit post (), the standard for V2G has finally been published. Most people don't realize that a bunch of hardware will need to be installed house side to use V2G. But we do! In theory, Tesla could add support for V2G protocols to the Backup Gateway and PW+ (since that's the...

teslamotorsclub.com

(by "I will direct" I mean that any further posts on that topic will be moved into one of those threads).

Thanks.


if only search worked better.

 

[Redhill_qik]

Here is the PG&E VPP Fleet Home data that I collected for the 9/7 event at 10 minute intervals. At 4:00pm when charging stopped it was at 842 kW, so not all Powerwalls were at 100% and more than the 742 kW on 9/6. The peak discharge was at 6:20pm at 25,279 kW with an estimated 49.4 MWh delivered during the event with 3,592 homes which are the new records.

Personally, I contributed 0 kWh to the grid as I was in an outage area (back up at 10:10pm), my Powerwalls did discharge 2.88 kWh against a baseline of 2.84 kWh so I may have made a whopping $0.08 if it counts. At least my Powerwalls will get back to 100% quickly today and I can then use solar to recharge my EV.

 

[jboy210]

A big question is will car makers warranty batteries that have been discharged/recharged all the time for these VPP events? I think there is always some battery degradation (it's no infinite), but cycling the batteries more as a daily thing will seem to increase warranty claims for car batteries, but no benefits for car makers at all.

It's one thing when there is a power outage and the Ford Lightning powers your home for a few days/a week, but another when a consumer is collecting a VPP check, and putting more use out of a battery that has an 8 year warranty. It's still early days so too early to see how this plays out.

That might be less of an issue as EVs shift to Iron-phosphate (LPF) batteries. These already allow charging to 100% more often and longer life. Supposedly CATL is working on a new higher capacity variant of these batteries in China.

 

[MJ_CA_2019]

Here is the PG&E VPP Fleet Home data that I collected for the 9/7 event at 10 minute intervals. At 4:00pm when charging stopped it was at 842 kW, so not all Powerwalls were at 100% and more than the 742 kW on 9/6. The peak discharge was at 6:20pm at 25,279 kW with an estimated 49.4 MWh delivered during the event with 3,592 homes which are the new records.

Personally, I contributed 0 kWh to the grid as I was in an outage area (back up at 10:10pm), my Powerwalls did discharge 2.88 kWh against a baseline of 2.84 kWh so I may have made a whopping $0.08 if it counts. At least my Powerwalls will get back to 100% quickly today and I can then use solar to recharge my EV.

View attachment 850293

This graph (and others you have shared) to me reinforces that the STORAGE capacity compared to the OUTPUT capacity of the PW2 is not optimized for grid support (or high load backup like AC use). I'd like to see 4x the continuous output rating in storage in future generations.

Ex. 5 kW continuous output should have 20 kWh storage.

 

[mjhwa]

That isn't how the VPP compensation works as it isn't how much is exported to the grid during the event hours from the Powerwall. The compensation is on the difference between the Powerwall discharge during the event hours and the baseline average for the same hours in the preceding days (10 weekdays or 4 weekend/holiday days).

If that's the case, there's no incentive for me to participate in VPP. VPP would only make sense is if someone had a large battery storage system such that Powerwall exports will exceed the solar export baseline average during those hours.

 

[cwied]

If that's the case, there's no incentive for me to participate in VPP. VPP would only make sense is if someone had a large battery storage system such that Powerwall exports will exceed the solar export baseline average during those hours.

I think you misunderstood the previous comment. It's just Powerwall discharge that is measured both to set the baseline and to calculate compensation. Whether it goes to the grid or to loads doesn't matter. That's why it's not necessary to look at the grid exports from Powerwall.

 

[Redhill_qik]

This graph (and others you have shared) to me reinforces that the STORAGE capacity compared to the OUTPUT capacity of the PW2 is not optimized for grid support (or high load backup like AC use). I'd like to see 4x the continuous output rating in storage in future generations.

Ex. 5 kW continuous output should have 20 kWh storage.

I think that the chart is mostly user choice related versus Powerwall specs. For instance my two Powerwalls can get me through 2:45 with a 5% reserve, I think the start of the roll off at the 1:30 mark is due to a much higher reserve setting of 30-50%. If we had the extra 6.5 kWh (20 kWh vs 13.5 kWh) that would meet the current user behavior or the user behavior could be incentivized to lower the reserve limit.

 

[mjhwa]

I think you misunderstood the previous comment. It's just Powerwall discharge that is measured both to set the baseline and to calculate compensation. Whether it goes to the grid or to loads doesn't matter. That's why it's not necessary to look at the grid exports from Powerwall.

Thanks for clarifying. Are you saying the compensation calculation is the difference of Powerwall discharge during VPP events and baseline average of the same hours during non-VPP events, i.e. discharging to my home?

Using Tesla's example on their website:

A one-Powerwall system with a Backup Reserve of 20% that is fully charged at the start of an event has 13.5 kWh * (100%-20%) = 10.8 kWh to contribute during that event.

If this system would typically serve 3 kWh of energy during the event hours, fully dispatching the Powerwall down to the Backup Reserve would deliver 7.8 kWh of additional kWh.

At $2 per kWh, this could be worth $15.60 for this event.

So in this example when it says "If this system would typically serve 3 kWh...", you mean that the 3 kWh is the discharge to the home?

 

[cwied]

Thanks for clarifying. Are you saying the compensation calculation is the difference of Powerwall discharge during VPP events and baseline average of the same hours during non-VPP events, i.e. discharging to my home?

Using Tesla's example on their website:


So in this example when it says "If this system would typically serve 3 kWh...", you mean that the 3 kWh is the discharge to the home?

Correct. I'm using "export everything" so I have Powerwall discharge to the grid as well, but if you don't turn that on then the discharge to the home is what sets your baseline. If you charge during those hours that might also affect the baseline, but since it's during peak that would be unlikely to happen.