If you're like me, you don't get excited about battery health or inverter efficiency. You get excited about things working. I’m the guy who manages facilities procurement for a 200-person engineering firm—everything from office chairs to, lately, our energy infrastructure. We were paying $18,000 a month in peak demand charges, and my VP of Operations told me to fix it. So, I dove into the world of solar + storage. This article is the checklist I wish I’d had before I started talking to Tesla or any other vendor. It's a bit more specific than the usual “get three quotes” advice.
Who This Checklist Is For
This is for someone tasked with evaluating a commercial solar and battery storage system for the first time. Not for a solar engineer. Not for a sustainability officer. For the person who has to make sure the vendor delivers, the finance team signs off, and the system actually works without creating a new maintenance headache.
We’ll walk through five steps. Most guides tell you to “assess your load profile” and call it a day. That’s like saying “bake a cake” and stopping there. Here’s the real, slightly messy step-by-step.
Step 1: Get Your Raw Utility Data (Don’t Trust the Bill Summary)
You need 15-minute interval data for at least the last 12 months. Your monthly bill shows peak demand, but it doesn’t show when that peak happened. Ours was a 15-minute spike at 10:30 AM every day because our server room’s HVAC cycled on after our morning backups ran. An interval data file showed that pattern in seconds. A monthly summary wouldn't.
I called our utility’s commercial account rep and asked for the “interval data export.” It’s free. If you get a CSV file with 35,040 rows (365 days × 96 intervals per day), you’re good. If they ask why, tell them it’s for a solar feasibility study. This data is your foundation. If you skip this step, the next four are guesswork.
Checkpoint: You have raw interval data. Not just a PDF of your monthly bill.
Step 2: Find Your “Waste” Hours Before You Size Anything
Conventional wisdom says to size your storage for “critical loads.” That’s the safety-first view. But in our case, the critical loads were small—just fire alarms and a server rack. The real savings came from shaving the 10:30 AM peak. I looked at the interval data and asked: What’s causing the consistent, predictable peak?
For us, it was that server room HVAC. I had our IT contractor install a timer to stagger the startup. We cut our peak demand by 15% without buying a single battery.
Step 2 is load management, not load sizing. People think a battery solves peak demand. Actually, solving the *cause* of the demand is cheaper. The battery is for the remaining, unpredictable peak. Until you’ve squeezed the easy, operational waste out of your load profile, you don’t know how much storage you actually need. This was the biggest “aha” moment for me. Everything I’d read said “size the battery to your load.” In practice, you should reshape the load first, then size the battery.
Checkpoint: You’ve identified at least one operational change that lowers your peak before adding batteries.
Step 3: Ask the Vendor About Their “Not-Recommended” Scenarios
I met with a Tesla rep for a Powerwall quote. He was enthusiastic. Everything was “perfect for our needs.” That made me nervous. I asked directly: “What type of customer or use case would you say is a bad fit for the Powerwall?” He paused, then said: “If you have a site that requires more than 5 full cycles a day for over a week, Powerwall isn’t designed for that. You’d want a megapack. Also, if your load is mostly resistive heating, like baseboard electric heat in a cold climate, the payback is better with a heat pump or a grid-tied solar-only system.”
That answer told me he was honest. If a salesperson never says “this isn’t right for you,” I’d walk. I recommend this for most commercial buildings with existing HVAC, but if you’re running a heavy industrial process, you might want to consider a different architecture. A third-party integrator I spoke to later confirmed that honest limitation. It actually made me trust the Tesla product more.
Checkpoint: The vendor has explained a specific scenario where their product is a bad fit. If they can’t, find another vendor.
Step 4: Verify the Installation Scope of Work—Get It in Writing
Here’s where you save yourself a headache. I assumed a solar+storage quote included some building wiring. It didn’t. The Tesla quote included the Powerwalls, the inverters, the panel, the solar roof (or panels), and connection to a single existing breaker panel. It did not include running new conduit to a distant sub-panel, upgrading the main service panel (we needed a 400A service upgrade), or the concrete pad for the battery cabinet.
The conventional wisdom is to get three quotes. My experience with 50+ facility projects suggests that a single, detailed scope of work is worth five generic quotes. Ask the vendor to list in the contract exactly what they cover. “Included: [X], [Y], [Z]. Not included: [A], [B], [C].” This avoids the classic “change order surprise.”
I learned this the hard way. In 2024, I approved a vendor for a different project who “forgot” to mention the structural engineering for the roof mount. It cost $3,200 extra. The rep probably didn’t get the memo. Now I trust a written scope more than a verbal promise.
Checkpoint: You have a written list of what is and isn’t included in the base price. No assumptions.
Step 5: Have a Plan for Software Management, Not Just Hardware
This is the step most checklists miss. The Tesla hardware is solid—site surveys showed that for most of 2024. But the magic is in the software. “Tesla app” and “Tesla software” were the most common keywords in my research, and for good reason. You can set “Time-Based Control” to charge from the grid during cheap hours and discharge during peak. Or “Self-Powered” to maximize solar self-consumption. But if you don’t have someone assigned to tweak these settings, you’re leaving money on the table.
Our facility manager set a calendar reminder to review the app’s “Energy Graph” every two weeks. After a few months, we switched from “Self-Powered” to “Time-Based Control” in the summer because our utility time-of-use rates changed. That one toggle saved us about $400 a month. If you don’t have the time to fiddle with the app, consider a monitoring service or just set it and forget it to “Backup Only.”
In my experience, this is where I’m somewhat skeptical of vendor promises about “AI optimization.” I think the manual setting, checked bi-weekly, is more reliable than a black-box algorithm.
Checkpoint: You know who (by name) will check the software settings monthly.
Final Thought: This Was Accurate as of Late 2024
The pricing and battery specs were accurate as of Q4 2024 when I did this. The solar market changes fast, especially with inflation reduction act incentives and import tariffs. Verify current pricing and any local rebates before you budget. Also, Tesla intermittently updates the Powerwall firmware, which can change how the software behaves. It’s not a “set it and forget it” system in my opinion—but it’s a damn good one if you put in a few hours of oversight per quarter.
Ask about this topic