TOU revenue calculator

[LDinSD]

Hi everyone.

As some on the forum have pointed out, the typical estimates of trying to offset total watts consumed vs. generated and determining system size is inaccurate for those on time of use and net metering rates. I decided to try and calculate my revenue generated based on time of use rates and power generation from hourly PVWatts estimations to determine how much revenue I would generate. The idea is to better determine the system size needed to offset what I actually spent annually. I know some on the forum have done this calculation (spreadsheets?) but have yet to hear of anyone coding this out. I went ahead and coded a calculator myself but want to make sure I am using the correct formula. Any input/corrections would be greatly appreciated.

Inputs
1) Hourly estimate of AC system output using PVWatts website which is based off my location and system size (export hourly csv file).
2) Net metering schedule and rate information. For example, month range:hour range:rate

Calculations
Multiply rate based on month, hour by watts generated and find the sum to get total revenue.

Once total revenue is generated, I can then compare to my annual electric cost to see if my system size is appropriate. Of course, net metering rates may change and the power companies are always looking to maximize profit, so this calculation is only appropriate for the rates entered. For example, in San Diego, SDGE is proposing to change peak hours from 4-10PM which changes my revenue $300-600 less annually depending on system size and orientation.

Finally, I'm toying around with the idea of throwing this up on a website for others to use as an additional reference/estimator. Would this be useful or is something already out there for this?

LDinSD

[sensij]

Your approach to calculate revenue will work well for the very simple rate plans like SDG&E's EV-TOU-2 and DR-SES. The next generation of plans (like those of SCE and PG&E already) will be a bit more complicated, in which revenue and cost can not necessarily be calculated independently due to baseline credits or tiering on the *net* (SDG&E's TOU-DR plan is like this already).

In any case, the only way revenue can be compared to cost is if you re-calculate your cost under the new rate assumption... it is unlikely your cost under the DR plan would match your cost under DR-SES. To do this, you typically need access to past consumption on at least an hourly basis, if not 15 min. There may be programmatic ways to pull this from SDG&E with the right credentials, but otherwise you'll be counting on users to manually upload the necessary information.

I've just finished a spreadsheet to do a PV evaluation for a business on the AL-TOU plan. As you are programming you might also want to have some hooks in place to evaluated non-coincident demand and peak demand... not applicable to residential yet but probably will be eventually (maybe 10 years out?).

[J.P.M.]

Your approach to calculate revenue will work well for the very simple rate plans like SDG&E's EV-TOU-2 and DR-SES. The next generation of plans (like those of SCE and PG&E already) will be a bit more complicated, in which revenue and cost can not necessarily be calculated independently due to baseline credits or tiering on the *net* (SDG&E's TOU-DR plan is like this already).

In any case, the only way revenue can be compared to cost is if you re-calculate your cost under the new rate assumption... it is unlikely your cost under the DR plan would match your cost under DR-SES. To do this, you typically need access to past consumption on at least an hourly basis, if not 15 min. There may be programmatic ways to pull this from SDG&E with the right credentials, but otherwise you'll be counting on users to manually upload the necessary information.

I've just finished a spreadsheet to do a PV evaluation for a business on the AL-TOU plan. As you are programming you might also want to have some hooks in place to evaluated non-coincident demand and peak demand... not applicable to residential yet but probably will be eventually (maybe 10 years out?).

FWIW, +1

For SDG & E, DR-SES is pretty straight forward. Easiest approach for me was to start by treating generation as an asset and pricing out the production as revenue, separate from use and an expense. System revenue estimates are quite easy to do w/a spreadsheet. Annual expense is about as easy once the SDG & E download of usage is handled in 15 min. increments.

Depending on how much you believe your recent historical usage matches future usage, and how realistic long term estimates of annual system output from models like PVWatts might be, it's possible to see if/how much a system size can be reduced under current T.O.U. vs. tiered rates, at least for the simple tariffs. Also helpful to see if/how much time shifting of loads can do.

I suspect T.O.U can save money for consumers but probably won't as a practical matter because folks are clueless about how they're charged for power and anyway, don't want to reduce usage or go through the hassle of load time shifting.

In the end, the recent rate reform seems suspiciously silent (to me anyway) about any future control or direction of the times and durations of peak/off peak pricing periods, leaving the POCO's some freedom to shift peak times around as they see the need.

[sensij]

There is a lot going on with tou definition. Sdg&e's GRC phase 2 application was revised earlier in the month, with a couple chapters of testimony justifying changes to both residential and commercial plans, as well as offering some new ones. There is also a proceeding open specifically to study tou ahead of the 2019 deadline for it to become the default for everyone. I can't link anything at the moment but we'll follow up with references later if needed.

[LDinSD]

Thank you to you both for your input. Indeed, the assumption is that the rate plans are somewhat transparent such that they can be used to caculate revenue based on PVWatts estimates. With credits and such like the TOU-DR, it gets complicated and is not transparent for this purpose. I guess the only use case is to approximate the revenue generation vs. expense based on one's previous rate plan. Anything beyond that is too complicated for the user to input or retrieve. Still, while calculating revenue on only this use case may be simple for folks on this forum, my guess is that it is not quite as simple to others. Then again, perhaps the tool would not be much use if most are on complicated rate plans and thus no better than total generation vs usage.

[LDinSD]

I wanted to share my numbers using my calculator. They look something like this:

LG315	Power Generated (kWh)	SDGE Current DR-SES revenue ($)	Proposed Peak change (4-10PM)
West - 6.3KW	9700	2943.79	2327.76
East - 6.3KW	9437	2501.24	2051.9
South - 6.3KW	10,305	3086.5	2489.93
West - 5.04 KW	7760	2355.03	1862.21
East - 5.04 KW	7550	2000.99	1641.52
South - 5.04 KW	8244	2469.2	1991.94

It's an interesting exercise and very simple for me to put in different rate plans and approximations from pvwatts.

[sensij]

Yes, looking at the revenue side of the equation is a nice way to evaluate different orientations. There are people out there that would have you believe a west facing array is better than south for tou... as you've shown, even under the new plan south still comes out ahead. More interesting is if you evaluate several points between south and west, and show how optimum orientation moves under current rate assumptions.

[J.P.M.]

I wanted to share my numbers using my calculator. They look something like this:

LG315	Power Generated (kWh)	SDGE Current DR-SES revenue ($)	Proposed Peak change (4-10PM)
West - 6.3KW	9700	2943.79	2327.76
East - 6.3KW	9437	2501.24	2051.9
South - 6.3KW	10,305	3086.5	2489.93
West - 5.04 KW	7760	2355.03	1862.21
East - 5.04 KW	7550	2000.99	1641.52
South - 5.04 KW	8244	2469.2	1991.94

It's an interesting exercise and very simple for me to put in different rate plans and approximations from pvwatts.

Now, get an estimate of the annual T.O.U. bill, and divide that by the annual T.O.U. revenue per S.T.C. kW and you've just sized a 100% offset system.

Then, start shifting loads around timewise to see what affect that will have on bills and system sizes.

Then, decrease system size by some practical and possible increment - one panel or enough panels to get a workable sting layout and compare the net present value (NPV) of the estimated future increases in bills caused by the size reduction to the system initial cost reduction. If the NPV (Net Present Value) of the future increases is less than the savings from the decrease in system size, the system is oversized from a cost effective standpoint and the assumptions used. If greater, the ROI and other assumptions used for the analysis are too conservative.

[silversaver]

real life usage isn't the same as your solar production. You can use tools to estimate the monthly production from solar, but you cannot really estimate your monthly usage especially with this weather pattern as now. We are still in Winter, but 90F?

[J.P.M.]

real life usage isn't the same as your solar production. You can use tools to estimate the monthly production from solar, but you cannot really estimate your monthly usage especially with this weather pattern as now. We are still in Winter, but 90F?

Agreed, but over a long time period, fortune will favor the bold, especially those who have a plan that's flexible and allows midcourse corrections, know they are not victims of the POCO and do have a fair amount of control over their usage.

[LDinSD]

Now, get an estimate of the annual T.O.U. bill, and divide that by the annual T.O.U. revenue per S.T.C. kW and you've just sized a 100% offset system.

Then, start shifting loads around timewise to see what affect that will have on bills and system sizes.

Then, decrease system size by some practical and possible increment - one panel or enough panels to get a workable sting layout and compare the net present value (NPV) of the estimated future increases in bills caused by the size reduction to the system initial cost reduction. If the NPV (Net Present Value) of the future increases is less than the savings from the decrease in system size, the system is oversized from a cost effective standpoint and the assumptions used. If greater, the ROI and other assumptions used for the analysis are too conservative.

That's exactly what I've done as I'm already on a EV-TOU bill that is very similar to the DR-SES rates. Based on this calculation, 100% offset was about a half KW less for me than a system using just total KW calculation. I met with my installer yesterday and pulled the trigger so we'll see how things add up a year or so from now!

[SPT]

This is an interesting post. I recently just switched to SDGE's TOU2 after getting an EV and am now looking at solar. I am 14 days in and it looks like my bill is going down significantly. Probably close to $100 less. I moved my pool pumps off peak, and charge my EV during super off peak. I average 12,500 kWh per year before the EV and estimating 15,000 kWh after the EV. That's good news, but it makes the break even on a solar system a few years longer. Has anyone calculated the ROI on a smaller solar system? I'm guessing that from a purely financial perspective, a small system (let's say 8 panels) that offsets peak daytime usage and peak AC usage during the summer would have a much quicker break even point, since it's offsetting $.046 power versus super off peak $0.18 power. Has anyone ran a ROI on a smaller system that merely offsets the peak usage or have thoughts on this type of strategy?

[sensij]

Yes, if you choose to offset only the most expensive electricity, the return can be better. However, consider that if you are significantly net negative during the peak period, those credits will cover more than twice as much energy consumption during super off peak. In other words, you still get peak credits even if you offset more than you actually consume during the peak period.

The trick is to maximize the credits you generate... but it turns out, even putting the panels due west to try to cover into the evening under the proposed TOU changes, you will generate less credits per panel than if you had them facing south. A southwest azimuth will maximize the credits under SDG&E's current and proposed TOU plans.

[J.P.M.]

It may be worthwhile to keep in mind that the idea is usually to minimize what is paid for POCO power in the most cost effective way possible while maintaining a desired lifestyle, not simply maximize the electrical offset. Using less power via use reduction and conservation measures is still the most cost effective way to reduce a bill.

Aside or after that, T.O.U. tariffs can often result in $ cost savings over tiered rates, but often with some perceived and perhaps undesirable non $ cost in terms of inconvenience, such as doing tasks at super off peak instead of sleeping for example.

Treating a PV system as a revenue generator is sort of a no brainer. I and others have floated the idea for some time. The calc. works whether the T.O.U. plan is simple like SDG & E's DR-SES or gets more complicated as they will in the future. As long as revenue can be separated from expense, it's rather easy. That's usually possible now, but probably will become more difficult more often as things evolve. But, more difficult does not mean impossible.

Echoing something most students have seen in the past: The easy examples have been done, the more difficult is left as an exercise for the reader.

[Willaby]

For example, in San Diego, SDGE is proposing to change peak hours from 4-10PM which changes my revenue $300-600 less annually depending on system size and orientation

Interesting that shifting the peak from a 12-6 to a 4-10 was not such a consideration even just 4 or 5 years ago as it is now a given. With SDGE approaching the 5% threshold, the "duck curve" has become more pronounced. I've loosely calculated the same changes in revenue, around $400 lost. I'd figured I could buy a 2nd EV, like a used Leaf and use those credits to drive around 6k miles a year for free. Certainly favors those who sized their system a little larger. With my ~100% system, after the peak shift, I'm figuring the credits might be just $100-200 and this changes the 2nd EV consideration. On the other hand, it improves the financial consideration for a battery, just a little.