Solar Lighting Night Shift

Five Spec Sheet Claims That Fail Basic Arithmetic

◦

After Dark | By Piotr Mikus

Browse the series: After Dark

(What to demand before accepting the performance claims on an all-in-one solar luminaire datasheet):

  • Lumens stated at the luminaire, measured through the lens and enclosure, with an LM-79 report and an IES file
  • Any “comparable to” HID claim backed by delivered lumens against the half-lumen guideline, not by wattage alone
  • Panel watts, battery watt-hours and battery voltage printed on the same sheet that makes the runtime claims
  • Worst-month daily energy harvest for the project site, with panel geometry, wiring and tilt stated
  • Battery chemistry named, with the manufacturer’s charging temperature floor and the battery management behaviour below it
  • Full-night energy in watt-hours for every programmed profile, including the motion-triggered state
  • Rated life of the LED, the battery cycles and the warranty stated on one consistent basis
  • The safety listings and ingress rating that apply where the product will be installed

A datasheet is a contract the manufacturer wrote about itself. Somebody still has to check it against the arithmetic.

A datasheet for a decorative all-in-one solar fixture reached my desk recently. It claimed comparability with a 150 W HID, a charge time in single-digit hours, dusk to dawn operation in several dimming modes and a rated life measured in decades. Every sentence was grammatically fine. I read them against arithmetic, standards and weather. Several fail before the weather is even considered. The rest fail when it arrives. A solar fixture sold on numbers like these costs the end user a dark pole and costs the whole industry its credibility.

A Comparison Claim Needs Lumens on Both Sides

Apparently a number printed on a sticker has been promoted to a lighting design.

A 150 W high pressure sodium lamp produces roughly 16,000 lumens initial and about 14,400 mean. Those are bare-lamp lumens, and a post top optic throws away a large share of them. That is why the working guideline for an LED replacement is about half the lamp figure, so 7,000 to 8,000 delivered lumens. A fixture that delivers a quarter of that or less is a different lighting level, whatever the comparison line says.

Two more things hide inside the lumen figure.

An efficacy near 100 lm/W is an LED array number. The light still has to leave through a polycarbonate or glass lantern and a decorative frame, and an enclosure loss in the range of 10 to 20 percent (my approximation) comes off the top. Without an LM-79 report, nobody can say what actually leaves the fixture.

The comparison line can also come from a different product class. A 150 W incandescent is about 2,600 lm. When a retail listing says 150 W replacement and the datasheet says HID, somebody changed the lamp type and kept the number.

Charge Time Is Division Until Somebody Checks the Sun

The sun has never been asked to comply with a datasheet.

Charge time claims are often battery watt-hours divided by panel watts. A 100 Wh pack and a 10 W panel gives 10 hours. That result requires the panel to deliver its full rated power for ten hours straight: 1,000 W/m2 of sunlight, cells at 25 C, the sun square to the surface and zero charge losses.

No day in the winter months in the central United States provides that. A well tilted south-facing panel at 39 degrees north collects something like 2 to 3 equivalent full-sun hours on a typical winter day, and less in a gray week. The pack then accepts that energy with losses. An empty pack takes days to refill with the lamp off, not hours.

When the documents for one product give three different charge times, none of them came from a test.

A Panel Wrapped Around the Top of the Fixture Sees Half the Sky at a Time

Rated watts count every cell as if the whole top of the fixture faced the sun.

A panel built into the faceted or conical top of the fixture presents cells to every compass direction. At 39 degrees north on 21 December the noon sun sits about 27 degrees above the horizon. A facet pitched near 40 degrees and facing south meets that sun almost square-on. The facets facing north get diffuse light only. East and west facets get a fraction of the beam.

In my own modelling of a fixture top of that shape, December daily energy came out near 70 percent of a south-facing tilted panel with the same cell area at the same site. In June the fixture top catches up and passes it. Winter is the design month.

Wiring can make it worse. If the cells share a single series string without bypass diodes per facet, the darkest facet limits the current and the output falls toward what the north side sees, roughly half of the reference in the same model.

A datasheet that prints panel watts and omits wiring and orientation has described the panel in a laboratory and said nothing about the pole.

Dusk to Dawn Is a Profile, and the Profile Needs a Budget

Programmable modes are easy to print and hard to power.

Add up the profile. A 20 W lamp that runs five hours at 40 percent and seven hours at 10 percent needs 5 x 8 W plus 7 x 2 W, so 54 Wh every night. A December harvest from a panel rated in single-digit watts is on the order of 10 to 20 Wh a day. The dimming modes are a schedule. The energy budget decides whether they ever run.

Motion boosts add load on top. A 30 second burst at 20 W costs 0.17 Wh. One hundred detections cost 17 Wh, which can exceed the entire day’s harvest. To pay for them, the baseline has to fall, often to a few percent of rated output. My earlier post, Full Output for Ten Seconds Is Not a Lighting Design, covers what that background state does to the photometrics.

A datasheet should state which profile its runtime claim assumes, how many detections per night, and what the controller does when the battery runs low. Without those three, the dimming modes are decoration.

Long Life, Cold Chemistry and Missing Listings

Rated life lives wherever the largest number fits.

An LED rated 100,000 hours is 22.8 years at twelve hours a night. A lithium battery rated for 1,500 cycles is 4.1 years at one cycle a day, and a starved pack cycles deeper, which shortens that further. The warranty tends to sit near the battery number. When one product carries two life ratings and a retail listing carries a third, the paperwork is telling you which one to believe.

Cold decides the rest. Lithium cells do not accept charge below 0 C (32 F) without plating lithium. A central US winter spends weeks with daytime highs near that line, so the charge window the energy budget assumed can be closed on exactly the days the budget needs it. A charging range printed up to 65 C (149 F) overshoots what most cells accept. The datasheet should say what the battery management system does at both ends of the range.

Then read the listings. CE and FCC marks do not make a luminaire a listed product in the United States. Ask for the safety listing and ingress rating that apply where the fixture will stand, and for the report behind each one.

Closing Thought

A datasheet is a promise written in the best season of the year. Physics does the audit.

If a claim cannot name the site, the month and the profile it was calculated for, treat it as a brochure. Every end user who buys a light on an inflated number and finds the pole dark by November ends up blaming solar lighting, and the whole industry pays for a figure one manufacturer chose to print.

Sources and Where to Verify

Piotr Mikus is a roadway lighting designer and specifier focused on solar powered street lighting and controls.

Continue reading the series: After Dark