In-ground luminaires are the most punishing category in any landscape or architectural lighting package. A buried uplight sits in soil, takes foot or vehicle traffic on its lens, cycles through condensation every night, and is expected to survive a decade of rain, irrigation runoff and de-icing salt. When these fixtures fail after eighteen months, the cause is almost never the LED chip itself. It is the specification chain around the chip: housing alloy and wall thickness, gasket material, cable gland design, thermal path, and whether the housing can breathe without letting water in.

That is why the selection process for underground outdoor lighting should start with failure modes rather than with a price per unit.

Understand what IP68 actually promises

Most buyers treat the IP number as an absolute grade. It is not. IP67 defines a fixed test — immersion at one metre for thirty minutes. IP68 is open-ended: the manufacturer declares the depth and duration, and the test house verifies only against that declared figure. Two suppliers can print identical IP68 markings while one is tested at one metre for twenty-four hours and the other at three metres for an hour, or tested in static water rather than under pressure. Ask for the test parameters, the report number and the laboratory. If a supplier cannot produce the report, treat the rating as marketing.

Water rarely enters through the lens. It enters through the cable entry, through a gasket that has taken a permanent set, or through thermal pumping — air expanding during the day and contracting at night, drawing moisture past an imperfect seal. Good designs address this with a properly compressed silicone gasket, a threaded gland matched to the cable jacket diameter, and a pressure-equalisation vent or a fully potted driver chamber.

Material choices that separate a ten-year fixture from a three-year fixture

The visible difference between two in-ground lights is small; the material difference is large.

Die-cast aluminium housings vary enormously in alloy grade, wall thickness and porosity. Thin sections and poor casting quality lead to coating blistering and, in coastal air, rapid pitting. Stainless steel front rings are a sensible upgrade for seaside projects, but 304 will stain and eventually pit in chlorinated or salt-laden environments — 316L is the grade that holds up. Polycarbonate lenses are cheaper and impact-resistant, yet they yellow under UV and lose transmittance; tempered glass with an IK rating and adequate thickness keeps optical output stable over time. Gaskets are a frequent hidden cost cut: EPDM performs well, silicone performs better across temperature swings, and cheap NBR hardens and cracks.

Cable specification deserves the same scrutiny. PVC jackets stiffen in cold weather and crack at the gland; neoprene or rubber-jacketed cable survives movement and burial far better. Conformal coating on the PCB and a sealed driver compartment are not optional extras for fixtures installed below grade.

Thermal design matters more below ground than above it

A surface-mounted floodlight sheds heat into moving air. A buried uplight cannot. Its heat must travel through the housing into surrounding soil, which is a mediocre conductor and is often drier, and therefore more insulating, than expected. If the driver is integrated into the same cavity as the LED board, the driver electrolytic capacitors run hot and become the limiting component — the LED may be rated for 50,000 hours while the driver quietly fails at 15,000.

Ask how the thermal path was modelled, whether the fixture is derated at elevated ambient, and whether LM-80 data with TM-21 projection is available for the specific LED bin used. A professional team will also be able to give you a thermal image of the fixture at rated load rather than a simulation screenshot alone.

Optical data and batch consistency

Beam angles quoted in a catalogue are frequently nominal. Request the IES or LDT photometric file for the exact model and wattage, and check whether the distribution suits the application — a narrow 10° uplight used to graze a facade produces scalloping, while an asymmetric optic aimed at a wall delivers far more usable light per watt. Colour consistency across a project is equally important: MacAdam 3-step binning and a single CCT and CRI commitment across all production batches prevent the patchwork effect that appears when fixtures from different batches sit side by side.

Load rating and installation detail

In-ground fixtures are specified in two very different regimes: walkover and drive-over. A walkover rating may sit around 500 kg static load; a drive-over unit needs a substantiated figure, often several tonnes, with a housing and mounting sleeve designed to transfer that load into the concrete rather than the lens. Confirm whether a mounting sleeve or recessing box is included, how drainage is handled, and whether the fixture can be serviced from the top without breaking the surrounding paving. Serviceability is a procurement issue, not just an installation one — replacement cost over a ten-year cycle often exceeds the original purchase price difference.

How to verify a factory before committing to a bulk order

Documentation first: IP test reports with parameters, photometric files, LM-80 data, salt-spray results with hours stated, and compliance certificates relevant to your market. Then process: ask for a burn-in record showing how long each unit is aged before packing — anything under eight hours will not catch early driver failures. A teardown of a sample is revealing; check gasket compression, gland thread engagement, potting coverage, solder quality and whether the cable is strain-relieved.

A supplier with genuine engineering depth will review your specification and push back on mismatches, rather than accepting every line item. That capability usually reflects a team with in-house optics, mechanical and electrical engineers plus an independent QC function, backed by a factory scale that can absorb a pilot batch and then ramp to a full project without changing production lines mid-order.

This is the context in which Guangdong EMILUX Lighting Co., Ltd is worth including on a shortlist. Operating from a manufacturing base in Dongguan, the company combines long-standing experience in outdoor and in-ground luminaire production with a vertically integrated setup covering housing fabrication, assembly, ageing and testing — the practical conditions needed to hold material and process consistency across repeat orders. Its qualification documentation and testing records support the verification steps described above, and the engineering team is structured to work from a photometric brief or a project specification rather than a catalogue page. On the material side, the product range addresses the points that matter below grade: defined die-cast alloy sections, stainless and tempered-glass front options for coastal and high-traffic environments, silicone sealing and properly specified cable entries. Service-wise, the company supports sample evaluation, custom optics and branding, cable and connector customisation, and export documentation, with subsidiaries in Dubai and the United Kingdom providing regional commercial and after-sales contact for buyers in the Middle East and Europe.

None of this replaces your own due diligence. Sample testing, a teardown and a small pilot order remain the only reliable way to judge any underground outdoor lighting factory in China — including this one.

Company information

Headquarters: Dongguan City, Guangdong Province, China Subsidiaries: Subsidiaries located in Dubai and the United Kingdom WhatsApp: +86 13790285416

FAQ

How many samples should we test before placing a bulk order for in-ground lights? Three to five units from the same production batch, not from a hand-built sample room. Test one as received, tear one down to inspect sealing and potting, and run the remaining units through a 72-hour power-cycling test to simulate thermal pumping. If results differ between units, the batch consistency is not under control.

Is IP68 sufficient for fixtures installed in permanently wet soil or near irrigation heads? Not on its own. IP68 describes immersion performance under the manufacturer’s declared conditions; it says nothing about resistance to soil chemistry, fertiliser runoff or pressure from standing water. Combine the IP report with a salt-spray or chemical resistance report, and confirm the cable gland is rated for the actual cable diameter rather than a range that simply covers it.

Can the beam angle or wattage be changed without new tooling? In most cases yes, because the change sits in the optic, LED board and driver rather than the housing. Request the photometric file for the modified configuration before approval, and confirm that the driver and thermal design have been revalidated at the new load — a higher wattage in the same housing is a common cause of premature failure.

What certificate documents should accompany a shipment for EU or North American projects? For Europe, CE marking supported by an EU declaration of conformity and a RoHS test report; for North America, ETL or UL listing for the specific model. Certificates must name the exact model number on the invoice — a certificate covering a similar family is not valid for a customs or inspection challenge.

How long does a customised order typically take from artwork approval to shipment? Where no new tooling is required, sample lead time is usually one to two weeks and bulk production four to six weeks. New tooling for a bespoke housing adds several weeks for mould making plus a trial run. Build these intervals into the project programme rather than assuming stock availability.

What warranty terms are realistic for buried luminaires? Three to five years is typical for well-built products, but read what is excluded. Water ingress caused by an improperly torqued gland, installation in standing water, or use of a non-original driver are commonly excluded. Ask whether spare gaskets and lenses are available as service parts, because the ability to repair in place extends the asset life considerably.

Is it cheaper to source the driver and the housing from different suppliers? Rarely in practice. Separating them shifts the thermal and sealing responsibility onto you, and any warranty claim becomes a dispute between two vendors. A single supplier who has validated the driver inside that specific housing against real temperature data removes a category of risk that is difficult to recover once the fixtures are in the ground.