Why Most LED TCO Calculations Are Wrong
If you’ve ever sat through a lighting sales pitch, you’ve seen the spreadsheet: “You’ll save $47,000 over 5 years!” Sometimes that’s accurate. Often it’s not. The gap between projected savings and real-world performance comes down to what gets left out of the calculation.
This guide walks through what actually affects the total cost of LED lighting over its lifetime — not the vendor’s version, but the facility manager’s version, where labor rates are real and downtime has consequences.
What TCO Actually Includes (and What Sales Decks Skip)
A proper TCO model has five cost buckets. Most vendor calculations include only two.
- Capital cost — Fixtures, drivers, controls, mounting hardware. What the invoice says on day one.
- Energy cost — The one vendors lead with. Usually accurate, sometimes optimistic on operating hours.
- Maintenance cost — Lamps, drivers, access equipment, electrician time. This is where projections go wrong.
- Disruption cost — Production downtime, after-hours premiums, occupied-space constraints. Rarely quantified, frequently significant.
- End-of-life cost — Removal, disposal, replacement. Often ignored because it’s 10+ years out. But it determines whether your “25-year LED” actually lasts 10.
If your TCO spreadsheet doesn’t have line items for all five, it’s incomplete.
Maintenance Cost: The Multiplier Everyone Underestimates
The single biggest cost advantage of LED over fluorescent or HID isn’t energy — it’s maintenance. But that advantage only shows up if you spec the right components and maintain them correctly.
Driver Failure: The Real LED Weak Point
LED chips rarely fail within 10 years. Drivers fail. Mean time between failures (MTBF) for quality drivers is 50,000–100,000 hours. For cheap drivers, it’s 15,000–30,000 hours. That difference shows up as truck rolls.
A single truck roll (electrician + lift + 45 minutes on-site) costs $200–$500 depending on access difficulty and union labor rates. If you have 500 fixtures and a 5% annual driver failure rate, that’s 25 truck rolls per year at $300 each = $7,500/year in maintenance labor alone.
Spec quality drivers. The $15 difference per fixture disappears against $300 truck rolls.
Group Relamping vs. Spot Replacement
Old-world thinking: replace failed lamps one at a time. LED world: plan group replacements at predetermined intervals because the cost of access exceeds the cost of the component.
For high-bay and hard-to-access locations, the economic choice is almost always group replacement of drivers at 60–70% of rated life, not waiting for failures. The math:
- Spot replacement: 1 failure × $350 truck roll = $350 each
- Group replacement: 20 drivers × $40 each + 1 day labor $800 = $1,600 total, or $80 per fixture including labor
Group replacement wins when access is difficult. The crossover point is roughly two spot replacements per year per 100 fixtures.
Energy: The Operating Hours Assumption That Inflates Savings
Vendor energy calculations assume 12–24 hours/day operation. That’s accurate for warehouses and exterior lighting. It’s wrong for offices (8–10 hours/day plus weekends = ~2,200 hours/year, not 4,400).
Use your actual utility billing data to determine operating hours. If you don’t know, here are realistic defaults:
- Office lighting: 2,000–2,500 hours/year
- Warehouse/industrial: 4,000–6,000 hours/year
- Exterior/security: 4,000 hours/year (dusk-to-dawn) or 12,000+ (24/7 security areas)
- Cold storage: 8,760 hours/year (lights on whenever the space is powered)
Every 1,000 hours of overestimated operating time inflates energy savings by roughly 30–40%. Keep the calculation honest.
Disruption Cost: The Invisible Expense
Disruption cost is what you pay when lighting work interferes with operations. It’s rarely itemized, but it’s often the largest single cost component in a retrofit.
Examples:
- Cold storage: Warming the space for safe worker entry = product temperature excursion risk + energy to re-cool. Some facilities won’t allow it, which means lighting can only be serviced during annual defrost/shutdown.
- Cleanroom/pharma: Any ceiling work requires re-certification of the space. $5,000–$20,000 depending on room class.
- 24/7 manufacturing: Lights must be serviced during planned downtime or shift change. Limited windows = premium labor rates.
- Retail/office: After-hours or weekend work. 1.5× to 2× standard labor rates.
When disruption cost is high, the TCO argument shifts dramatically toward higher-spec (more reliable) fixtures, even at higher capital cost. A $20/fixture premium for better ingress protection or a higher-quality driver pays for itself the first time you avoid an unplanned shutdown.
Fixture Life: Why “50,000 Hours” Doesn’t Mean 50,000 Hours
LED lifetime is usually quoted as L70 @ 25°C ambient — the point where lumen output has decayed to 70% of initial. Two problems with using this number naively:
- Ambient temperature: Every 10°C above 25°C halves LED life. A fixture rated for 50,000 hours at 25°C might deliver 25,000 hours at 35°C (common in high-bay installations with poor air circulation).
- Driver life: Drivers fail before LEDs. A fixture is only as reliable as its weakest driver. L70 tells you nothing about driver MTBF.
Real-world specification approach: assume useful fixture life is 70% of the L70 rating in controlled conditions, and plan maintenance intervals accordingly. For critical applications, design for group replacement at 60% of rated life.
Warranty: What’s Actually Covered (and What’s Not)
Warranty terms vary wildly. A “5-year warranty” can mean very different things:
- Repair or replace at manufacturer discretion — The standard. They’ll ship you a new driver or fixture. You pay shipping and labor.
- Includes labor — Rare, valuable. Removes the largest cost component of warranty claims.
- Pro-rated — After year 3, you pay a percentage of replacement cost. Better than nothing, but adds administrative overhead to track.
- On-site swap — Premium tier. Manufacturer sends a tech. Extremely valuable for difficult-access locations.
For large projects (>500 fixtures), negotiate a stocking agreement: manufacturer pre-positions 2–5% spare fixtures or drivers on-site. Cuts warranty response time from weeks to hours.
End-of-Life: Planning the Next Replacement
LED fixtures don’t suddenly “burn out” like fluorescent lamps. They fade. Deciding when to replace becomes a judgment call rather than a clear failure event.
Practical end-of-life planning:
- Budget for replacement at 70% of L70 life — By then, light output has decayed noticeably and the risk of component failure is rising steeply.
- Don’t wait for a failure cascade — When drivers start failing, they tend to fail in clusters (same batch, same operating hours). Once failure rate exceeds 5%/year, plan a group replacement.
- Factor in technology change — The fixtures you install today will look obsolete in 10 years. Plan for a full refresh, not just component replacement, after 10–12 years.
Set aside a sinking fund: $X per fixture per year into a replacement reserve. When year 10 arrives, the capital is there and the decision is financial, not emergency.
TCO Calculation Template: The 5 Numbers You Need
You can build a proper TCO model in Excel with five inputs per scenario. Here’s the structure:
Sum these over your analysis period (typically 5 or 10 years). Compare scenarios on total cost per fixture-year, not just upfront cost or just energy savings.
The Payback Reality Check
Simple payback (capital ÷ annual savings) is the metric everyone uses and the metric that misses the most important factors. A 2-year payback sounds great until you realize it assumes 12-hour daily operation in a building that’s only occupied 8 hours a day.
Better metrics:
- Net present value (NPV) over 10 years at your organization’s discount rate. Accounts for the time value of money and gives a clearer picture than simple payback.
- Internal rate of return (IRR) — The discount rate at which NPV = 0. Useful for comparing lighting projects against other capital investments.
- Sensitivity analysis — What happens to payback if energy prices rise 20%? If operating hours are 20% lower than assumed? If maintenance costs are 2× projected? The scenarios where the project still makes sense are the ones worth doing.
Run the numbers honestly. If the project still works with conservative assumptions, it’s a good investment. If it only works with aggressive savings projections, it’s a risk.