10kVA vs 20kVA UPS, 36V LiFePO4 vs Lead-Acid: A Buyer's Comparison From Someone Who's Wasted $38,000 on Wrong Orders

The Comparison I Wish Someone Had Given Me in 2019

I handle power backup orders for commercial projects. Eleven years now. I've personally signed off on roughly 300 UPS and battery orders, and I've made about 14 significant mistakes that cost my company around $38,000 total. Not proud of that number.

Here's what I've learned: the UPS buying decision breaks down into three comparisons that actually matter. Not the spec sheet comparisons — the ones that cost you money when you get them wrong.

• 10kVA vs 20kVA UPS — capacity, obviously, but the real difference is in how you'll load it
• 36V LiFePO4 vs lead-acid batteries — upfront cost vs five-year reality
• Router/modem UPS vs "real" UPS systems — the philosophical divide that cost me a $200 order and a $20,000 relationship

Let me walk through each one.

Comparison 1: 10kVA vs 20kVA UPS — The Load Curve Trap

On paper, this looks simple. 20kVA carries twice the load. Done. But that's not where the money goes wrong.

In 2021, I ordered a 10kVA UPS for a client's network operations center. Calculated load: 7.8kVA. Comfortable margin. What I didn't calculate: the client added two additional server racks six months later. Suddenly we're at 9.4kVA running on a 10kVA unit.

Result: The UPS ran at 94% capacity for four months. Efficiency dropped, heat output spiked, and the batteries aged about twice as fast as they should have. That "saved" $4,100 on the smaller unit cost us $6,800 in premature battery replacement and an emergency upgrade later. Total damage: roughly $10,900.

Here's the counterintuitive part — or rather, the causation most people get backwards: People think higher kVA means better reliability. Actually, it's load matching that determines reliability, not raw capacity. A 20kVA UPS running at 25% load is often less efficient than a 10kVA UPS running at 50% load. You want to sit in that 40–70% sweet spot.

My working rule now: Order the UPS that gives you 40–60% load at current demand, then verify the unit supports parallel expansion for later. If it doesn't support parallel expansion, go one size up.

When 10kVA makes sense

• Stable load forecast — you know the equipment won't change for 3+ years
• Space-constrained installs (20kVA units need serious floor space and cooling)
• Budget-constrained projects where load matching can be maintained

When 20kVA makes sense

• Any sign the facility will grow (and they almost always grow)
• Critical loads where redundancy matters more than efficiency
• When the unit supports parallel configuration — future expansion without replacement

Comparison 2: 36V LiFePO4 vs Lead-Acid — The Five-Year Math

This one makes people angry. (Myself included, for a long time.)

Lead-acid is cheaper upfront. A 36V 100Ah LiFePO4 battery pack might run $1,200–1,800. Equivalent lead-acid: $400–600. Clear winner, right?

No. Wait — let me rephrase that. That's the comparison that keeps people buying lead-acid, and it's the one that cost my company $7,400 in a single year.

Here's what actually happened on a 2022 project. We spec'd lead-acid for a 36V UPS system at a remote telecom site. Upfront cost: $520 for the battery bank. Over the next 26 months:

• Battery replacement at month 22: $520 (plus $380 in truck rolls and labor)
• Reduced capacity from month 14 onward — the site had two brief outages that the lead-acid bank couldn't fully bridge
• Environmental control (ventilation for hydrogen off-gassing) added $1,100 to the install

LiFePO4 equivalent total: $1,400 upfront, zero replacement, zero ventilation cost, deeper usable depth of discharge (80–90% vs 50% for lead-acid).

Over five years, the lead-acid path cost roughly $3,100 per site. The LiFePO4 path: $1,400. That's before factoring the outage costs.

Industry data backs this up. Per IEC 62619:2022 (the safety standard for industrial lithium batteries), LiFePO4 cells maintain 80% capacity after 3,000–5,000 cycles. Lead-acid: 300–500 cycles at 50% depth of discharge. That's not a small gap — it's a 10x maintenance difference.

The real trade-off: LiFePO4 costs more today, but the crossover happens around month 18–24. If your project horizon is under two years, lead-acid can pencil out. Past that, you're paying a premium for the "cheaper" option.

(Note to self: I really should build a simple cost-per-month spreadsheet for the team. We keep having this conversation.)

Comparison 3: Router UPS vs Commercial UPS — The Small Order I Almost Lost

This one's personal.

In March 2023, a small IT services company — two guys, maybe five clients — called about a UPS for a Wi-Fi router and modem setup. Small order. Maybe $180. My first instinct was to pass it to a junior rep and move on to the 20kVA quote on my desk.

I took the call anyway. Took 20 minutes. Spec'd them a small line-interactive UPS with a 36V LiFePO4 pack — overkill for a router, honestly, but it gave them 4+ hours of runtime on their network edge gear. (Side note: a router UPS is not the same as a desktop UPS. Router loads are low-draw, continuous — you want runtime, not surge capacity.)

Last month, that two-person company placed an order for a 20kVA UPS system for a client's server room. Referral. $23,000 order.

Small doesn't mean unimportant — it means potential. That's not a motivational poster line. It's just math. The vendor relationships I still have from my early career started with orders under $500. The vendors who treated those orders seriously are the ones I still use for $50,000 orders.

I've seen the other side too. In 2020, we needed an urgent 36V LiFePO4 battery for a test deployment. Four units. A distributor's rep literally told me "we usually don't do orders under 50 units." Cool. I've spent roughly $340,000 with their competitor since then.

What small buyers actually need

If you're ordering a UPS for Wi-Fi router backup, modem/router combo, or a small home office:

• Runtime over capacity. A 36V 20Ah LiFePO4 pack will run a typical router + modem for 6–10 hours. A 600VA UPS will run it for 45 minutes. Size for hours, not VA.
• Line-interactive is probably enough. Unless you're in an area with severe voltage fluctuations, double-conversion online UPS is overkill for a router.
• Battery chemistry matters more than brand. LiFePO4 at 36V will outlast lead-acid by years — even in small formats.

So Which One Do You Actually Need?

Depends on what you're protecting. Here's my honest breakdown:

If you're spec'ing a 10kVA vs 20kVA UPS for a facility: Calculate your load, add 30% headroom, and check whether the unit supports parallel expansion. If growth is possible, go 20kVA. If the load is truly stable for 3+ years, 10kVA at 50–60% load is more efficient. Don't size up for bragging rights — size up for real growth or redundancy needs.

If you're choosing between 36V LiFePO4 and lead-acid: Unless your project horizon is under 24 months, LiFePO4 wins on total cost of ownership. The upfront premium pays back around month 18–24 through cycle life, depth of discharge, and eliminated maintenance. For permanent installs, it's not close.

If you're a small buyer looking at a UPS for Wi-Fi router or modem backup: Don't let anyone make you feel like your order doesn't matter. Size for runtime, pick LiFePO4 if the budget allows, and don't overthink the brand. The suppliers who answer your questions seriously today are the ones you'll call for the big orders later.

My experience here is based on about 300 commercial power backup orders, mostly in the $2,000–$50,000 range. If you're working with larger infrastructure — data center scale, utility-grade — your experience will differ significantly. But the three comparison traps above? Those show up at every scale. They've cost me $38,000 to learn. Hopefully this saves you a chunk of that.