The Solar Journey

Why we went solar

The interest predates the house. The budget and the reason didn't arrive until 2020 and 2021 — and by WB's own account, two separate decisions (heat pumps, then solar) came out of two separate conversations.

The idea didn't start with a sales call. In January 2017, working at GE Intelligent Platforms, WB drafted a private pilot proposal of his own — he called it “Predix Solar Project v1.1” — to monitor a solar installation on GE's Predix industrial platform. He'd already been running what he called “Predix in a box” at home: Raspberry Pi, Intel NUC, and Edison boards doing his own home-based monitoring, the same instinct wherewhen and wattwhen both descend from. That's where the want came from — solar generation and battery storage, monitored the way he was already monitoring everything else at home — four years before the Bowmans had the roof, the budget, or a reason to act on it.

The budget arrived through Mass Save. The Bowmans' first 0% HEAT Loan closed sometime in March 2020 — $17,000, a figure that only surfaces in the record two years later, when CLEAResult's Michael Polletta cited the program's $25,000-per-property combined cap while processing a second loan. By WB's own account, it was that direct conversation with Mass Save — not the solar shopping still to come — that settled the household on heat pumps. Some of that same work paid for an electrical upgrade: when Ace's own site inspector asked about the panel a year later, Dalio Electric's Thomas D'Alio confirmed it was already a 200-amp main lug.

Then COVID moved WB's desk into an unheated, uncooled basement. The Bowmans redid the space so he could work there year-round and added Mitsubishi mini-splits — by his own account, mid-2021. With the whole household home 24/7 and the majority of heating and cooling now running through those mini-splits instead of oil or gas, the electric bill climbed to around $400 a month. By WB's own math, financing the solar system would run close to $500 a month — close enough to what they were already handing the utility that, at that point, they figured they'd be paying themselves instead.

Batteries specifically were justified by outages, not just the bill. The 2011 “Snowtober” nor'easter — power out for the better part of a week across Worcester County, the same National Grid territory Mendon sits in (see “Why batteries” below) — was still fresh enough in memory that it shaped the decision on its own: without the well pump running, there's no water pressure, so no flushing toilets, and no air conditioning or heat either, whatever the season. Solar panels alone don't fix any of that — a grid-tied inverter shuts itself down the moment the grid does, for line-worker safety, same as every other home's. Batteries were the piece that actually kept the house livable through an outage; the panels just kept the electric bill down on a normal day.

The specific push to actually go solar, by WB's own account, was Solarize Mendon-Upton — the town's run of the state's Solarize Mass Plus program. Its “Ask the Experts” webinars, held April 8 and 13, 2021 (solarizemendonupton.com — the only URL on that site that still resolves), covered solar PV, battery storage, EVs, heat pumps, geothermal, and solar hot water. Mendon had already invested in its own municipal solar farm to offset the town's own energy costs — part of why the town ran the program at all. The Bowmans went with heat pumps and solar. Solarize's installer list is also the direct reason Ace ended up on the quote list: Ace's Nick Moore, Mar 2, 2021 — “Thanks for your time today, glad you found us through the Solarize program and reached out!”

Who else got a look

Six installers, one town program, five weeks — Feb 24 to Mar 31, 2021.

What set Ace apart, before anything went wrong

It wasn't just Tesla. By WB's own account, every installer on the list, Ace included, started out quoting the same way: standard 330W panels spread across both roof faces — south (rear) and north (front) — to maximize panel count, the front's poor exposure notwithstanding. Tesla's quote is the one that survives in writing, because WB put the question to them directly, Mar 8, 2021: “What size system do we get with just the panels on the back of the house only. As the front of the house is north facing.” Tesla's ZD answered the next day: pull the front panels and the rear-only array drops to 29 panels, 9.86kW — a real number on what the north roof was actually contributing.

What actually set Ace apart was being the first installer willing to accept larger-wattage panels — 395W LG modules instead of the 330W everyone else defaulted to — concentrated on the good, south-facing roof alone. It cost more per panel, but by WB's own math it cost less than the extra 330W panels a both-faces design would have needed on the marginal north side to make up the difference. The system spec below is the result: all 34 panels at 26° tilt, 174° azimuth — nearly due south. None on the north face at all.

Tesla stayed in the running longer than the panel math alone would explain, because of the battery: Powerwall was the name in home storage at the time, ahead of the LG RESU10 Ace was quoting. What killed it was Tesla's own website showing install dates more than three years out, with — by WB's own account — no real confidence Tesla would install in Massachusetts at all. WB cancelled Mar 10, 2021: “We are not looking for solar panels on the front of our house, and with the lower power generation would not offset the cost enough to justify... we will be working with other companies to do only the back side of the house with large capacity solar panels.”

What was quoted, and what got installed

The signed Master Purchase Agreement, the signed design page, and ACE Solar's own Aurora shading/irradiance report, next to solar_system_summary.md's later inventory of what's actually on the roof and in the garage — one table instead of two, so the two sides sit side by side.

Item	Quoted — Mar 2021 contract	Installed — as of Aug 2026
System size	13.43 kW DC PV array	13.43 kW DC total
Modules	(34) LG395Q1C-A6, 395 W	34 × LG395Q1C-A6, 395 W each
Inverter	(1) SolarEdge SE7.6kW EnergyHub — 25-year warranty included, as contracted	SolarEdge SE7600H-USSNBBL14 Energy Hub, 7.6 kW AC (one string)
Optimizers	(34) SolarEdge P505 Power Optimizer — the contracted part number (Schedule A)	34 × SolarEdge P401, one per module — a lower-tier part; see “What went wrong” below
Battery	(2) LG Chem RESU10	2 × LG RESU10H Prime, 9.6 kWh usable each — ~19.2 kWh nominal, ~10 kW combined discharge. Installed in the garage.
Monitor / racking / electrical / labor / permits	Included in scope, per the signed contract	—
Roof	—	26° tilt, 174° azimuth — nearly due south
Energized	—	June 21, 2021 (SolarEdge site record, Site ID 2428857)
Programs	—	MA SMART SMANG_40299 · National Grid ConnectedSolutions storage rewards
Monitoring today	—	mySolarEdge / SolarEdge Monitoring — the ACE Solar owner app was discontinued Oct 2025
Total price	$55,086.00 — $37,486 solar PV + $17,600 batteries	—

SolarEdge Monitoring, Aug 2026.

The signed design page — 15,276 kWh estimated annual production.

Aurora Shade Report, Apr 9 2021 — 92%/84% annual solar access, 87.8% weighted TSRF, ACE Solar's own signed certification. The one tree the report flagged near array 2 was later trimmed, then cut down entirely.

Schedule A of the signed contract — the P505 optimizer line, in Ace's own letterhead.

LG RESU10H Prime datasheet — the manufacturer's own recommended 59–86°F charge window, against a garage installation.

$/kW means average kW, not peak kW

National Grid's ConnectedSolutions battery incentive isn't a flat payment for how many kW of battery sits in the garage — it's paid on the average kW the system actually delivered across the events it was dispatched for that season, not the battery's peak or nameplate rating. A battery that discharges hard for ten minutes and then drops out — which is exactly what Ace's inverter kept doing mid-event, see the timeline above — drags the average down, and the payment down with it, even though the hardware itself is rated far higher.

WB found this out running his own numbers, not from anything Ace or National Grid explained upfront. Emailing Ace's CM in May 2023 about how much he'd actually been paid: “Technically it is high as I have the $ per KW at 275 (Cell S30). The real rate is $225 which matches the payment I received of $203.” His own tracking spreadsheet's naive $/kW cell said $275; the number that actually matched a real payment was $225 — per average kW delivered across the season's events, not per kW of battery installed.

That's the same kind of number in the stat blocks below — 8.3 kW average delivered across 12 events in 2023, not the batteries' roughly 10 kW combined discharge rating. It's also exactly what wattwhen, this whole page's reason for existing, is built to track for itself: an independent, event-by-event read of what actually got delivered, instead of trusting a dashboard summary or waiting months for a utility's own true-up.

The other side of the ledger nobody puts a number on: when the batteries recharge from the grid instead of from solar — topping off ahead of a scheduled event, or overnight — that's paid-for electricity going back into the battery, not free sunlight, and it's a real cost against whatever ConnectedSolutions pays out. It stays invisible unless something is actually watching which source charged the battery each time — the same reason wattwhen cross-checks two independent meters instead of trusting either one alone.

What a bad outage around here actually looks like

By WB's own accounting, batteries pay for themselves three ways: keeping the house running through an outage — not just a week-long disaster, but the far more common few-hours-to-a-day kind too — shifting evening and nighttime usage onto stored solar instead of the grid (see the scenario table above), and Connected Solutions event payments from National Grid (the “ConnectedSolutions math” section above). Outages are the hardest of the three to put hard evidence behind, and the one with the most riding on it.

National Grid's own outage map only shows what's happening right now — there's no public tool to look up outage history by address. These two are pulled from published storm coverage and the state's own emergency-management outage map instead, the closest public evidence of what the sub panel above is actually for.

Which is also the honest caveat: the same battery backup was already installed for the Dec 2023 storm, and — as the rest of this page documents — it's the one job it's failed every time it's been asked to do. The value the batteries were sold on is real; whether this particular install ever delivered it is a separate question.

What Ace and SolarEdge said, next to each other

Every quote below is verbatim from the email or support-case record, with the speaker's initials, role, and the date — full names are in the "Who's who" key at the foot of the page. Nothing is paraphrased.

Aug 2026

The inverter Ace scheduled in July 2023 is in. Ace's Residential division closed in 2025 and its owner app is gone; SolarEdge's mySolarEdge is the primary manufacturer channel now, kept alive by a $47.80 cellular-plan renewal in July 2026. The system runs, but for five years the only account of whether it was running well belonged to whichever vendor answered the phone that week. wattwhen exists so that stops being true — SolarEdge's numbers cross-checked against an independent Emporia meter on the same interconnection, not one company's word taken on faith.