Forty panels, 16.1 kW DC capacity, three EG4 6000XP inverters, and eleven ECO-LFP48100 battery packs providing 99.44 kWh of storage. My solar system went live in late April 2025 and I've been tracking every single kilowatt-hour. Let me show you exactly what this system does to a real electric bill in coastal Georgia.
Not a simulated bill. Not a solar company's estimate. Not projected savings from a proposal. The actual invoice. Every month.
The System
Forty panels, 16.1 kW DC capacity, three EG4 6000XP inverters, and eleven ECO-LFP48100 battery packs providing 99.44 kWh of storage. The system went live in late April 2025. I call it the Spartina Solar system because that's what Home Assistant calls it, and at this point Home Assistant is basically the household's fourth occupant.
The house is on the Georgia coast — Midway, just south of Savannah. If you're not familiar with coastal Georgia: it's hot, humid, and the AC runs from May through October without apology. Summer electric bills without solar were genuinely uncomfortable.
The House
The loads are worth naming because they matter for context. Two electric water heaters. No pool. A Tesla Model Y (Augustus) and a Volvo XC60 PHEV (Eunice), both charging at home. Eunice is almost always topped off by solar during the day — plug her in when we get home and the panels handle the rest. Augustus charges at night, two or three times a week, during the cheap super off-peak window.
The Rate Structure: Know What You're Playing With
My utility — Coastal Electric Cooperative — uses time-of-use (TOU) pricing. Before you can understand what's happening in these bills, you need to understand the rate structure, because it's the whole game.
| Pricing Tier | Hours | Rate (per kWh) |
|---|---|---|
| On-Peak | 4 PM – 8 PM | $0.20 |
| Off-Peak | 6 AM – 4 PM & 8 PM – 10 PM | $0.095 |
| Super Off-Peak | 10 PM – 6 AM | $0.05 |
| Fixed Facilities Charge | — | $33/month |
That rate spread is the most important number in this whole story. Peak electricity costs four times what super off-peak electricity costs. If you're charging two EVs and running a household battery bank with 99.44 kWh of capacity, and you're willing to automate around that spread, you can do a lot of damage to your bill.
On top of the per-kWh rates, there's a Power Cost Adjustment (PCA) — a per-kWh adder that fluctuates monthly and isn't prominently advertised on the bill. More on that in its own section, because it deserves one.
"The rate spread between peak and off-peak is where battery storage makes financial sense. The battery acts as a time-shifting device, using stored energy during expensive hours and charging during cheap hours."
The Pre-Solar Baseline: Three Months of Honest Numbers
I have three months of clean pre-solar TOU data from the same billing system: February, March, and April 2025. Solar came online late April, so these three months represent what the house consumed before the panels changed anything.
| Month | Days | Total kWh | Bill |
|---|---|---|---|
| Feb 2025 | 37 | 2,333 | $294 |
| Mar 2025 | 28 | 1,471 | $199 |
| Apr 2025 | 31 | 1,597 | $210 |
| AVERAGE | — | 1,800 | $234 |
Three months is a limited pre-solar dataset — I'll acknowledge that. It doesn't include the brutal coastal Georgia summer, which historically pushes bills to $300+. But it's the only pre-solar TOU data I have with tier-level detail from this billing system, so $234/month is the honest baseline I'm working from.
When I say solar cut my bill by 56%, I mean from $234 pre-solar to $103 post-solar. Not from an estimated number. Not from a simulated baseline.
The Full Month-by-Month Table: All 14 Months
| Month | Days | Total kWh | Bill | Notes |
|---|---|---|---|---|
| Feb 2025 | 37 | 2,333 | $294 | Pre-solar |
| Mar 2025 | 28 | 1,471 | $199 | Pre-solar |
| Apr 2025 | 31 | 1,597 | $210 | Pre-solar, solar activated late Apr |
| May 2025 | 30 | 622 | $109 | First solar bill — partial month |
| Jun 2025 | 31 | 957 | $145 | First full solar month |
| Jul 2025 | 30 | 1,050 | $125 | |
| Aug 2025 ⚠️ | 35 | 1,665 ⚠️ | $161 | Billing software change; data unreliable |
| Sep 2025 | 26 | 694 | $117 | Includes $34 past-due from Aug; current = $83 |
| Oct 2025 | 30 | 118 | $44 | Zero on-peak usage — best month |
| Nov 2025 | 30 | 552 | $83 | Zero on-peak usage |
| Dec 2025 | 31 | 995 | $124 | |
| Jan 2026 | 31 | 1,713 | $154 | Heavy EV charging month |
| Feb 2026 | 31 | 1,494 | $142 | Heavy EV charging month |
| Mar 2026 | 28 | 788 | $90 | Most recent month |
Post-solar average (10 months, excluding August): $102.80/month
The Peak Collapse: From 251 kWh to 27 kWh
This is the number that tells you the most about what the battery system is actually doing.
Pre-solar, the house averaged 251 kWh of on-peak consumption per month — electricity bought from the grid during the 4-8 PM window at $0.20/kWh. That's 251 kWh every month getting charged at four times the cheapest rate of the day.
Post-solar, the 10-month average on-peak consumption (excluding August) is 27 kWh. Most of that 27 kWh average is pulled up by June and July 2025, when the system was still in its early months and I was dialing in the battery management automations. Look at what happens as the system matures:
- June 2025: 162 kWh on-peak
- July 2025: 34 kWh on-peak
- September 2025: 6 kWh on-peak
- October 2025: 0 kWh on-peak
- November 2025: 0 kWh on-peak
- December 2025: 1 kWh on-peak
- January 2026: 1 kWh on-peak
- February 2026: 1 kWh on-peak
- March 2026: 9 kWh on-peak
October and November: zero. December, January, February: essentially zero — one kilowatt-hour each, probably a battery dispatch edge case. The system is almost entirely eliminating on-peak grid purchases. That's the battery doing its job.
"Going from 251 kWh on-peak to near zero isn't just a billing win — it's proof that the TOU arbitrage is working the way it's supposed to. The battery system isn't just providing backup power. It's systematically routing consumption away from the most expensive hours of the day."
What's happening mechanically: the 99.44 kWh battery bank charges during the day on solar production, or overnight at the super off-peak rate if solar isn't available. By 4 PM when peak pricing kicks in, the batteries are full. The house runs on stored energy from 4 to 8 PM. The grid sees almost nothing.
The $44 October Bill: Understanding the Floor
October 2025 produced a $44 electric bill. Total consumption: 118 kWh for the entire month. On-peak usage: zero kilowatt-hours.
Break that down: $33 fixed facilities charge + $11 in actual energy charges. Eleven dollars of electricity for a month. In a house with two EVs, two water heaters, and regular HVAC.
October is the convergence point. Temperatures have dropped enough that the AC stops running around the clock. Solar production is still solid — shorter days than summer, but still good hours in coastal Georgia. The batteries hit 100% most days, which means the system is exporting excess solar or at minimum eliminating virtually all grid purchases. And on-peak consumption is zero, so that $0.20/kWh rate applies to exactly nothing.
$44 is the floor. It's $33 + minimum energy charges, and it demonstrates the theoretical ceiling of what this system can achieve when conditions are favorable. I'm not suggesting October is representative of the average — it's clearly the best month in the dataset by a wide margin. But it's useful calibration: it shows you what "working perfectly" looks like, and that informs your expectations for the other eleven months of the year.
The EV Charging Strategy: Super Off-Peak Is the Play
January and February 2026 are the two highest-consumption months in the post-solar dataset. January: 1,713 kWh total, with 1,497 kWh in the super off-peak tier. February: 1,494 kWh total, with 1,308 kWh super off-peak.
Those super off-peak numbers aren't the house load. The house doesn't consume 1,497 kWh between 10 PM and 6 AM. That's primarily Augustus charging overnight two or three times a week and the battery bank topping off when solar production was short during the day.
The Math
| Scenario | Rate | Cost for 1,497 kWh |
|---|---|---|
| Super Off-Peak (overnight) | $0.05 | $74.85 |
| Off-Peak (daytime) | $0.095 | $142.22 |
| On-Peak (evening) | $0.20 | $299.40 |
Super off-peak electricity costs $0.05/kWh. In January, I consumed 1,497 kWh at $0.05 — that's $74.85 for all that overnight energy. At the off-peak rate ($0.095), the same 1,497 kWh would have cost $142.22. At on-peak ($0.20), it would have been $299.40.
But the comparison I actually care about is gas vs. electricity for the Model Y. Augustus, the Tesla Model Y, gets roughly 3.5-4 miles per kWh in real-world coastal Georgia driving. At $0.05/kWh overnight, I'm fueling the car at an equivalent cost of roughly 1.2-1.4 cents per mile. Gas in coastal Georgia right now runs around $3.75/gallon. A comparable 30 MPG gas car costs about 12.5 cents per mile. The EV at super off-peak rates costs roughly 89% less per mile to fuel.
The Automation
The automation that enables this is automation.spartina_smart_grid_charge_v2_2 running in Home Assistant. Every evening it checks the next-day weather forecast and decides how much grid charging to request for the battery bank: if solar production is forecast to be strong, it charges less from the grid overnight and lets the panels fill the bank the next day; if it's a cloudy week, it tops the batteries at $0.05/kWh while everyone's asleep. Augustus's charging windows are set for super off-peak (10 PM-6 AM, $0.05). He charges two or three times a week — not every night. Eunice charges during the day on solar whenever she's plugged in. On-peak charging for either car simply doesn't happen.
The PCA Nobody Talks About
Every bill in this dataset carries a Power Cost Adjustment — a per-kWh adder on top of the base rates. The utility adjusts it monthly based on their fuel and power purchase costs, and it fluctuates without much fanfare.
| Period | PCA Rate (per kWh) |
|---|---|
| Most months (Feb 2025 – Mar 2026) | $0.00985 |
| Jul 2025, Jan 2026 | $0.00785 |
| Dec 2025 | $0.01285 |
At the standard $0.00985 rate on a 2,333 kWh bill (February 2025), the PCA adds roughly $23. On the October 2025 bill with 118 kWh, it adds about $1.16. The PCA is a real cost that scales directly with consumption.
In December 2025, the PCA jumped to $0.01285 — the highest in the dataset. On 995 kWh of consumption, that's $12.79 in PCA. It's not catastrophic, but it's also not zero, and it's not a number that appears prominently in most discussions of TOU rate structures.
Solar doesn't eliminate the PCA, but it does reduce the kWh base it's applied to. When October consumption drops to 118 kWh, the PCA contribution drops proportionally. One more reason total consumption reduction matters, not just shifting when you consume.
The August 2025 Anomaly: Being Straight With You
August 2025 data is not reliable, and I'm not going to pretend otherwise.
The billing period ran from July 15 to August 19, 2025 — a 35-day window. During this period, Coastal Electric changed billing software and could not extract actual interval usage data. The result was an adjusted bill: the $161 charge appears to be a reasonable approximation, but the kWh figures and TOU tier breakdown that appear on that invoice do not reflect what actually happened in the house during that period. The numbers on the bill — 1,665 kWh total, 1,351 super off-peak, 19 on-peak — are artifacts of the billing adjustment, not real meter reads.
I've excluded August 2025 from every aggregate calculation in this article. The $161 bill amount itself is real (I paid it), but I can't tell you the TOU breakdown was real.
The Bottom Line: $234 to $103
| Metric | Pre-Solar (3 mo avg) | Post-Solar (10 mo avg) | Savings |
|---|---|---|---|
| Avg Monthly Bill | $234 | $103 | 56% |
| Total Monthly Consumption | 1,800 kWh | 879 kWh | 51% |
| On-Peak Consumption | 251 kWh | 27 kWh | 89% |
Fourteen months. Three pre-solar, eleven post-solar (one of those eleven with unreliable kWh data but a real bill amount).
Pre-solar average (Feb-Apr 2025): $234/month, 1,800 kWh, 251 kWh on-peak.
Post-solar average (May 2025 – Mar 2026, excluding Aug): $102.80/month, 879 kWh, 27 kWh on-peak.
That's a 56% reduction in average monthly bill, a 51% reduction in total consumption, and a 89% reduction in on-peak purchases. With two EVs charging at home throughout.
The Math on the Investment
System cost: approximately $29,000. Federal Investment Tax Credit at 30%: $8,700. Net cost after ITC: $20,300.
| Item | Cost |
|---|---|
| 16.16 kW DC solar array | $4,250 |
| ~99 kWh battery bank (DIY) | $19,376 |
| Inverters, racking, wiring, BOS | $5,374 |
| Total before credit | $29,000 |
| Federal tax credit (30%) | -$8,700 |
| Net system cost | $20,300 |
Monthly savings: $131/month ($234 pre-solar - $103 post-solar).
At $131/month in savings: $20,300 ÷ $131 = 155 months, or about 12.9 years to payback.
I'll be honest about the uncertainty in that number. My pre-solar baseline is three months — winter/spring months that don't include a coastal Georgia summer, which historically runs $300+ per month for this house. The three-month average of $234 may understate the true annual average. If pre-solar bills averaged $260/month instead of $234, the monthly savings jump to $157 and payback drops to about 10.8 years.
The battery investment ($19,376 of the total $29,000) is the honest variable. Batteries don't just save money — they enable the on-peak elimination you see in the data, they keep the house running during grid outages (coastal Georgia gets hurricanes), and they make the EV charging strategy possible at scale. But they're also the most expensive component and have a finite lifespan. The financial case for the battery bank depends heavily on whether you value TOU arbitrage, outage resilience, and EV charging optimization. For this household, all three matter.
One Year In: What the Spreadsheet Says
Fourteen months ago, I had a house that cost $234/month to run on electricity, including two cars. Now I have a house that costs $103/month, including the same two cars. The system cut total consumption in half and nearly eliminated purchases at the most expensive pricing tier.
The sales pitch version of this story would show you the October bill — $44, wow, look at that — and call it typical. It's not typical. It's the best case. The honest version shows you all fourteen months in a table with TOU breakdowns, flags the anomalous August data, notes that January was $154, and lets you form your own conclusion.
Here's mine: the system does what it's supposed to do. It produces electricity during the day and stores the excess. It discharges during peak hours so the house doesn't buy expensive grid power at 4 PM. It charges the EVs cheaply overnight. The batteries made the difference between a solar installation that passively offsets consumption and one that actively manages cost.
$20,300 net of tax credit. $131/month in savings. Roughly 13 years to payback. A system that should run 25+ years. The math is not slam-dunk obvious, but it's not crazy either — and that's with a three-month pre-solar baseline that probably understates the summer savings.
The spreadsheet doesn't lie. It just requires more patience than the brochure.