Test Reports & Certification — Ballast Bracket

Sources

Who ran each test.

Each of these proves something different, and none of them replaces a project-specific structural check.

Accredited third party

UV ageing of the finished bracket

Tianjin Bailun Testing Technology Co., Ltd. Tested the complete bracket, not a material sample. Commissioned by us, not by the factory.

International third party

Thermal cycling of the tank body

TÜV SÜD Certification and Testing (China), Shanghai. Ageing and temperature cycling on the HDPE base tank, referenced to IEC 61215-2:2016 MQT 11. Three samples, no visual defects.

Production test rig

Uplift, cold storage, evaporation

Run on the line's own test rig, not at an accredited laboratory. Useful engineering evidence of how the part behaves — we do not present it as a certification.

Summary

The full list.

Test	Standard or method	Condition	Result	Laboratory
UV ageing	GB/T 16422.3-2022, method A, cycle 1	UVA-340 lamp, 8 h at 0.76 W/m²/nm and 60 °C, then 4 h condensation at 50 °C. 84 cycles.	Pass	Tianjin Bailun, CMA / CNAS
Thermal cycling	IEC 61215-2:2016 MQT 11, referenced	200 cycles between −40 ± 2 °C and +85 ± 2 °C, each cycle not exceeding 6 h.	Pass	TÜV SÜD, Shanghai
Dry heat	IEC 61215-2:2016 MQT 11, referenced	200 h at +105 °C, followed by visual inspection of the full surface.	Pass	TÜV SÜD, Shanghai
Cold storage	Plant procedure, −40 °C chamber	168 h and 336 h at −40 °C, three tank generations, weighed before and after.	Pass	Production test rig
Wind uplift	Pull-out rig, 1,111 N per clamp	Horizontal pull on both clamps, vertical to the horizontal plane; repeated at 60 °C substrate.	Pass	Production test rig
Evaporation	Outdoor mass tracking, live installation	One filled unit weighed repeatedly over 13 months, 2024-01 to 2025-02.	0.25 kg lost	Production test rig

These figures are the summary. The reports themselves are not published on this site and cannot be downloaded from it — full copies go to named buyers against a project enquiry, and the issuing laboratory can be contacted directly to verify any of them.

Detail

How each test was run.

UV ageing

Tianjin Bailun Testing
CMA · CNAS · ilac-MRA
Issued 2026-10-12

Sample: Plastic water ballast bracket, current production
Standard: GB/T 16422.3-2022 — plastics, exposure to laboratory light sources, fluorescent UV lamps. Method A, cycle type 1.
Lamp: UVA-340
Irradiance: 0.76 W/(m²·nm) at 340 nm
Exposure step: 8 h, black panel 60 ± 3 °C
Condensation step: 4 h, black panel 50 ± 3 °C
Duration: 84 cycles of 12 h = 1,000 h
Acceptance: No cracking, chalking or significant colour change after exposure
Result: Pass. No cracking, no chalking, no significant colour change.

This is the test that answers the first question a flat-roof buyer asks: will the plastic go brittle and crack in the sun. 1,000 hours under UVA-340 at 60 °C is an accelerated dose, not a calendar guarantee — it is evidence, and it is why the body is specified with a UV stabiliser package rather than plain HDPE.

Thermal cycling and dry heat

TÜV SÜD, Shanghai
Three samples, no visual defects

Sample: HDPE base tank, three pieces
Reference: IEC 61215-2:2016, MQT 11
Cold: −40 °C for 168 h
Hot: +105 °C for 200 h
Cycling: 200 cycles between −40 ± 2 °C and +85 ± 2 °C, no cycle longer than 6 h
Inspection: Full visual inspection of every sample after each programme
Result: Pass. No major visual defects recorded on any of the three samples.

Cold is the harder half of the problem for a water-filled plastic part, and it is why the working range is quoted at −40 °C to +85 °C rather than a comfortable European average. The report is on the tank body, not on the assembled bracket.

Cold storage at −40 °C

Production test rig
Cold chamber, two durations

Samples: Three tank variants in current production
Condition: −40 °C, two durations: 168 h and 336 h
Checks: Mass before and after, surface appearance, fill port and seams
Result: Pass. Mass stable, no cracking at the fill port or along the weld seam, water-tight after return to room temperature.

Run to cover cold-climate enquiries — northern China, Mongolia, Kazakhstan, northern Japan. The part is not rated for ice expansion inside a full tank; this test confirms the empty shell survives storage and transport at −40 °C without cracking.

Wind uplift

Production test rig
Pull-out rig

Method: Horizontal pull applied to each of the two clamps, perpendicular to the horizontal plane, i.e. the direction the wind lifts a module
Load: At least 1,111 N per clamp
Repeat: Run again with the substrate heated to 60 °C, the condition under which the plastic is softest
Reference: 50-year return wind pressure across Chinese provinces; most locations fall at or below 0.55 kN/m²
Result: Pass at 0.55 kN/m². No tearing or failure at either clamp, cold or at 60 °C.

Read this the right way: it is the clamp and tank resisting uplift at a given pressure, not a statement that your roof structure is adequate. Roof load capacity is signed off by the project structural engineer — see the limits below.

Evaporation over 13 months

Production test rig
Outdoor, on a live roof

Sample: One filled bracket, weighed on a fixed schedule
Start mass: 46.65 kg
End mass: 46.40 kg after 13 months
Total loss: 0.25 kg — 0.54 % of the water carried
Ambient: Weekly maximum between 8 °C and 32 °C; weekly minimum between 1 °C and 27 °C

For the question that always comes up — won't the water just evaporate — half a percent over a full year on a sealed moulded body means one top-up at most, and in most climates none. It was measured on a real unit outdoors across a full annual cycle, not estimated.

Field evidence

Two typhoons went over arrays our customers installed with this bracket.

These are dates on which completed arrays carrying this bracket met real storm loading and were still standing afterwards. We supplied the mounting system; the installation was done by the customer and their own crew, on their own roof, under their own engineer's sign-off.

Ballasted array installed by a customer on an industrial flat roof — A customer's array on an industrial roof. The bracket and the ballast layout are ours; the roof survey, the installation crew and the structural sign-off were theirs.

Date	Event	Location	What it means
2024-09-16	Typhoon Bebinca, force 14	Shanghai, Suzhou	Customer-installed ballasted arrays on industrial roofs came through the landfall wind field with no module loss and no displacement.
2023-07-30	Typhoon Doksuri, force 16	Fujian	Super typhoon-strength wind over a coastal installation built by the customer's crew; array intact afterwards.

Who did what, precisely. These roofs were installed by our customers, not by us. We supplied the bracket and the ballast layout; the roof survey, the installation labour and the structural sign-off were theirs. That is exactly why we cite them: the array held up in someone else's hands, on a roof we never touched. We do not present them as our own contracting work, and they are not a guarantee for your site — design pressure for your project comes from your local code and your own engineer.

Limits

What these tests do not prove.

0.55 kN/m² is the tested uplift resistance at the clamps. It is not the load-bearing capacity of your roof, and it must not be presented to a client as one. Ballast weight, point loads and membrane condition on an existing roof are signed off by the project structural engineer — we supply hardware, we do not take over that signature.

Four further limits

Results apply to the samples tested. Every report carries that limitation in its own wording. They are evidence about a production design, not a certificate for an arbitrary batch.
Three of the six programmes were run on a production test rig, not by an accredited third party. They are engineering evidence. Where independence matters, quote the UV ageing and TÜV SÜD results only.
No test here covers an installation method different from the specified one — full fill depth, clamp torque, module size range and row spacing all change the load path.
Accelerated hours are not years. 1,000 h of UVA-340 is a dose used to compare materials and to catch gross failures, not a calendar translation into a warranty period.

What was tested, and what came back.