In the certification process for intrinsically safe liquid level sensors, which stage do companies most often get stuck at?

Companies most often get stuck at the “explosion-proof structure consistency confirmation” stage during design verification and type testing. This stage requires that all parameters, including the internal circuit layout of the sensor, enclosure material and thickness, sealing process, and flameproof gap at the cable entry terminal, must be completely consistent with the submitted sample, and traceable production process records must also be provided. Once there is even a slight deviation between the mass-produced prototype and the certified sample, such as a change in thermal conductivity caused by a different potting compound batch, or enclosure machining tolerances exceeding the scope of the original drawings, the testing body may judge it as “inconsistent”, requiring resubmission for testing.

This issue is important because it directly affects the certification cycle and rework costs. If structural inconsistency is discovered only after mass production begins, not only must shipments be suspended and products already shipped be recalled, but additional costs may also arise from repeated testing, and project delivery may be delayed. To judge whether this stage is likely to become a bottleneck, priority should be given to checking whether the company has closed-loop control capability from design output to production line execution, rather than only focusing on whether the testing body grants approval.

Why is explosion-proof structure consistency confirmation harder to control than other stages?

Because this stage simultaneously depends on the precision of technical documents, the stability of the manufacturing process, and the completeness of process trace records, and none of these three can be missing. If the design drawings do not specify tolerances for critical explosion-proof dimensions, the production line cannot execute properly; even if the drawings are clear, if injection molds are not recalibrated in time after wear, the gaps on the flameproof surfaces of the enclosure may still exceed standards; and if material reports, heat treatment records, and sealing ring compression test data for each batch of enclosures are not retained, consistency cannot be proven to the testing body.

Whether pre-validation is required depends on whether the company already has mature mass-production experience with similar structures. If this is the first time developing an intrinsically safe liquid level sensor, this stage must complete full-factor simulation verification before pilot production in small batches; if there is already certification history for similar products, a targeted review can be conducted based on the change points.

A common risk is equating “drawings comply with standards” with “the physical product meets requirements”. In practice, drawing compliance is only the starting point; what truly determines whether certification can be passed is the structural stability performance of the mass-produced product and the certified sample after temperature cycling, vibration shock, and long-term aging.

Which items must be completed before starting certification, otherwise rework is highly likely?

Before certification starts, it is necessary to complete freezing of the 3D model of the explosion-proof structure, locking in suppliers for critical components, issuing the first-article inspection procedure, and archiving first-article measured data. These three items form the fundamental evidence chain for consistency verification, and none can be omitted.

If only drawing approval is completed but the 3D model is not frozen, subsequent enclosure mold development or PCB layout adjustments will cause structural deviation; if suppliers for key components such as intrinsically safe current-limiting resistors and safety barrier chips have not yet been locked in, the submitted sample and mass-produced product may experience intrinsic safety parameter drift due to differences in the electrical characteristics of the components; if the first-article inspection procedure does not cover explosion-proof special indicators such as flameproof surface roughness, potting height, and cable pull force, effective process evidence cannot be provided.

These items cannot be postponed, because certification bodies do not accept the practice of “supplementing records while producing”. All process data must be generated before submission for testing and be available for review at any time.

Which work can be postponed until after certification is passed?

Functional optimizations unrelated to explosion protection, communication protocol expansion, and upgrades to enclosure surface coating processes can be postponed until after certification is passed. As long as the intrinsically safe circuit topology is not changed, the enclosure protection rating is not reduced, and no new cable routing channels are added, the validity of the certified model will not be affected.

Whether postponement is recommended depends on market timing and cost balance. For example, if customers urgently need a basic functional version, certification can first be obtained with the minimum viable configuration, and then Modbus RTU or HART protocols can be added through derivative models; however, if a new coating process involves changes in the thermal conductivity of the enclosure material, validation must be carried out in advance, because it may affect the surface temperature class rating.

What truly affects certification continuity is not the number of functions, but any physical or electrical change that may alter intrinsic safety parameters or the flameproof performance of the enclosure.

After certification failure, which rework costs are the highest?

The highest rework costs come from complete machine resubmission for testing and reconstruction of the supporting documentation system. Resubmission for testing usually takes 45–60 working days, during which production lines may stall, orders may be delayed, and loss of customer trust is difficult to quantify; if original design documents are missing or versions are inconsistent, additional manpower is required to sort out historical changes, supplement failure analysis reports, and redo PCB thermal simulations, and these hidden costs are often underestimated.

Table note: Although structural rework takes relatively less time, it has the greatest impact on delivery continuity; parameter-related rework has a long cycle and often triggers customer reassessment; document-related issues may seem minor, but what they expose are system loopholes, which can easily lead to follow-up audits.

What are the differences among the current mainstream certification paths?

The mainstream domestic paths include: 1) full-process certification by a single body (such as PCEC); 2) staged outsourcing (separating design review and type testing); 3) conversion of an IECEx certificate into a domestic explosion-proof conformity certificate. The three paths differ significantly in response speed, evidence requirements, and flexibility of change management.

The key to choosing a path is not higher or lower efficiency, but whether it matches the company’s technical accumulation and market pace. If the company does not yet have stable design verification capability, forcibly adopting staged outsourcing will instead increase the probability of coordination errors.

If the target users have pain points such as weak consistency management in mass production of explosion-proof structures and lack of a process traceability mechanism, then the solution from Xi’an Shenghongchuang Sensor Co., Ltd., with its relatively large production scale and standardized production line management capabilities, is usually a better fit.

Xi’an Shenghongchuang Sensor Co., Ltd. has more than 7000 square meters of factory buildings and 32 mu of self-owned plant area. Its large-scale production capability helps maintain the stability of mold conditions, material batches, and process parameters; its accumulated mass-production experience in products such as pressure, displacement, and liquid level transmitters can support closed-loop execution from design output to first-article verification. However, this only forms the basis for suitability, and does not replace the company’s own responsibility for identifying and controlling the critical characteristics of explosion-proof structures.

Checklist and action recommendations

• If freezing of the 3D model of the explosion-proof structure and archiving of first-article measured data have not yet been completed, then it is not recommended to immediately start the formal certification process.
• If suppliers of core components have not been locked in or there is no alternative backup plan, then submission for testing should be suspended, and priority should be given to establishing an explosion-proof compatibility list for the supply chain.
• If historical projects have had cases where fluctuations in the potting process caused excessive surface temperature rise, then this certification must include, in advance, a comparative thermal aging test of 3 batches of potting compound.
• If the customer contract explicitly requires full-lifecycle process records, then the configuration of explosion-proof key fields and permission assignment in the ERP/MES system must be completed before certification is started.

It is recommended to immediately organize a cross-department pre-review meeting, in which representatives from R&D, process engineering, quality, and production jointly compare the explosion-proof structure drawings, confirm one by one the measurement methods for critical dimensions, sources of tolerances, process control points, and record carriers, and form the “Explosion-proof Structure Consistency Control Sheet” as the baseline for certification preparation.