The problem I keep seeing
I still remember walking a dusty fence line outside Hermosillo in May 2021 and seeing a bank of cells humming but not doing the useful work I expected—this is the kind of moment that sharpens my thinking about battery storage utility scale projects. In one scenario I managed, a 100 MW Li‑ion BESS sat idle during evening peaks even though it had the capacity; the system cut peak charges by 18% when it worked—so why do we keep building stacks that miss value when it counts? That sentence lays out the scene, the number, and the sting (claro, amigo).
I’ve been on the B2B side of supply for over 15 years, and I’ve seen the same design sins: oversized inverters that never run at rated load, control logic that ignores actual state of charge (SOC), and so-called “smart” software that treats forecast errors like a surprise party. Those choices aren’t just academic — in Sonora in 2021, a mis-specified inverter trimmed expected revenue by about $120k in the first quarter after commissioning. I’ll be blunt: vendors pitch bells and whistles, but the real user pain is predictable—poor ramping during ancillary services, hidden thermal constraints, maintenance windows that weren’t planned for. We call the problem out by name: systems built for specs, not for the messy, hot, real grid. — Next, let’s look at how I’d fix that.
From fixes to forward strategy (what I actually do)
Now I switch gears: I map problems into measurable fixes. When I lead a deployment, I start with three concrete checks—real measured round‑trip efficiency, actual inverter throughput under high ambient temps, and control‑loop behavior at low SOC. In 2020 I recommended a firmware tweak and a modest hardware change on a mid‑capacity project in Baja; within 90 days the asset was delivering peak shaving and frequency response reliably (and netted a 12% uplift in dispatch revenue). That’s the kind of comparative, forward-looking work that stops theory from bleeding money.
What’s Next?
We should compare options by outcomes, not features. For new bids I run scenario sims: high solar, cloud events, and multi-day low‑irradiance stretches. Then I push vendors to show how their BESS will behave across those runs — not just a single snapshot. You want proof of consistent SOC control, inverter thermal derating curves, and verified ancillary services performance. I’m asking for hard logs, not glossy decks. (No kidding.)
Here are three practical metrics I insist on when choosing a solution: 1) Measured availability during peak windows (not theoretical uptime), 2) Demonstrated energy throughput at specified ambient temps, and 3) Proven revenue capture for ancillary services over at least a six‑month period. Use those, and you’ll see real differences in ROI, dispatch reliability, and operational headaches. I say this from projects where the right tweak boosted net revenue by tens of thousands of dollars in months — small changes, big impacto. We learn fast, adapt faster. Stop buying promises; buy measured performance. — For the next steps, I’ll outline a test protocol and vendor acceptance checklist.
Final note: I keep giving this same advice in client meetings from Ciudad Juárez to Guadalajara, and it’s become a habit: prioritize the operational truth over the spec sheet. If you want a partner who’ll push for those logs and cross‑check the math, I’m that person — and I work with teams who know how to get it done. For solid, real‑world utility outcomes, consider practical implementation first, brand second. sungrow
