Przebieg r20261030-0348-v001-zadymiarka-concept-1

Regulatory plan — EU, US/Canada, ESTA/ANSI (ZAD-CRT-01, ZAD-CRT-02, ZAD-CRT-03, ZAD-SAF-08)

Rev A, 2026-10-30. Product: 950 W fog machine, concept C2 (see deliverables/02-concept/design-basis.md): EU 230 V and US 120 V variants, Class I, built-in 2.4 GHz BLE module (ESP32-C3-MINI-1 class), DMX512-A/RDM, handheld remote with 2×AAA, sold in EU and US.

Labelling rule: every number is SOURCE (url, accessed 2026-10-30), CALCULATION or ASSUMPTION (reason). A standard number without a year means "edition to confirm" (not verified). Costs are ASSUMPTION (typical 2025/26 lab price levels from the author's general knowledge; no lab quote read; get 2-3 quotes) unless marked SOURCE.

0. Findings that change the design or the plan (read first)

1. Metal enclosure vs. built-in antenna. The concept uses a powder-coated aluminium chassis with the ESP32-C3-MINI-1 class module (PCB antenna). A PCB antenna inside a closed metal box will not radiate; range and the module's modular grant are both at risk. Options: (a) plastic window / antenna area outside the metal (module at the edge, keep-out per module datasheet); (b) a -1U (U.FL) module variant with an external antenna: then only the antennas listed in that module's grant are allowed (KDB 996369 D03), otherwise a permissive change/new grant is needed. ASSUMPTION: the grant of the chosen module lists its antenna(s); to confirm from the module's FCC grant and RED/ETSI reports. Decide before prototype B.
2. Remote with 2×AAA is a separate radio product: own RED declaration, own FCC SDoC/ID labelling (the module's grant covers the module, not the remote host), own battery-regulation duties.
3. Safety standard route. UL 998 for the US (below, SOURCE). For the EU the plan uses EN 60335-1 with the applicable Part 2 (EN 60335-2-98 humidifiers is the nearest known; edition and applicability to confirm with the lab); manufacturers of comparable machines declare EN 60335-1 (SOURCE: Martin JEM ZR35 spec as quoted in search result, https://www.portlightingstore.com/products/martin-jem-zr35-1500-watt-professional-fog-machine, accessed 2026-10-30).
4. EN 60335-1 harmonisation status. SOURCE (secondary): the harmonised reference under the LVD is EN 60335-1:2012 amended up to A15:2021 (A11, A13, A1, A14, A2, A15:2021, AC:2014); EN IEC 60335-1:2023 is not yet cited; Implementing Decision (EU) 2025/1457 restricts part of cl. 20.2; the CENELEC amendment EN IEC 60335-1:2023/FprAB:2026 is in preparation (https://standards.iteh.ai/catalog/standards/clc/098e2071-55cd-403c-b4d2-6c6689c4156f/en-iec-60335-1-2023-fprab-2026, accessed 2026-10-30) and the IEC 60335-1:2020+AMD1:2025 adds Annex U (cyber security of remote communication) (https://webstore.iec.ch/en/publication/110266, accessed 2026-10-30). Consequence: test to the harmonised EN 60335-1:2012+A15:2021 for presumption of conformity, and ask the lab for a gap check against EN IEC 60335-1:2023 (our unit has BLE/USB-C service port, so Annex U / RED Art. 3.3(d,e,f) cyber duties must be looked at, see 1.1). The citation status must be re-checked in the OJEU list before the DoC is signed (edition to confirm).
5. Cyber-security of radio equipment. Delegated Regulation (EU) 2022/30 activates RED Art. 3.3(d),(e),(f) for internet-connected radio equipment and those handling personal data; applies from 1 Aug 2025 (author's knowledge, edition/standard EN 18031-1/-2/-3 to confirm). Our BLE remote link is not internet-connected, no personal data: ASSUMPTION that 3.3(d-f) are not triggered; document the reasoning in the technical file and confirm with the notified-body-or-lab opinion. Firmware update over USB-C: signed images recommended.

1. EU route

1.1 Legal framework

Route: RED 2014/53/EU (radio built in). Art. 3.1(a) health and safety (includes the LVD safety objectives, no separate LVD declaration), Art. 3.1(b) EMC (no separate EMC Directive declaration), Art. 3.2 efficient radio spectrum use. Module reports can be reused, but the final product still needs its own Art. 3.1(a), 3.1(b) and 3.2 evidence (radio tests on the host with the module installed; ASSUMPTION: partial reuse of the module's EN 300 328 results allowed if antenna, power and installation equal the module's report). Conformity assessment: Annex II (internal production control), possible because harmonised standards are applied in full; if EN 300 328 / EN 301 489 are not applied in full, a notified body is required (RED Annex III/IV). Cost of choosing a notified body is not in the totals.

1.2 Harmonised standards (edition status)

Cost columns are ASSUMPTION (EUR, accredited lab in PL/DE/CZ, 1 unit, 2 variants where noted). Pre-compliance = in-house or hourly rental (1 day of chamber/bench). Timeline counts from sample delivery.

Notes: (a) for each standard, the harmonised citation and edition must be checked in the OJEU lists for RED and LVD/EMC just before the DoC is signed ("edition to confirm" above, nothing in this table has been verified against the OJEU except EN 60335-1); (b) the sample and the test plan are agreed in a lab kick-off, including Part 2 selection and operating modes (idle, heating, pumping at 60 ml/min, DMX active, BLE active).

1.3 Draft EU Declaration of Conformity (RED)

EU DECLARATION OF CONFORMITY (No. ZAD-DoC-001, DRAFT)
1. Product: Fog machine "<model>" (EU variant 230 V ~ 50 Hz, 960 W), batch/serial no.: <...>
   Remote control "<model R>" (2×AAA, 2.4 GHz BLE) - a separate declaration number ZAD-DoC-002
2. Manufacturer: <name>, <address>, Poland, <e-mail>  [Authorised representative: n/a]
3. This declaration of conformity is issued under the sole responsibility of the manufacturer.
4. Object of the declaration: as in 1 (traceability: photo/ serial).
5. The object described above is in conformity with the relevant Union harmonisation legislation:
   Directive 2014/53/EU (RED); Directive 2011/65/EU (RoHS) as amended by (EU) 2015/863.
6. Harmonised standards and technical specifications applied (editions to be filled from the
   OJEU list on the signing date):
   EN 60335-1:2012 + A11, A13, A1, A14, A2, A15:2021 [+ Part 2 to confirm]; EN 62233 [ed.];
   EN IEC 55014-1 [ed.]; EN IEC 55014-2 [ed.]; EN IEC 61000-3-2 [ed.]; EN 61000-3-3 [ed.];
   EN 300 328 [V.]; EN 301 489-1 [V.]; EN 301 489-17 [V.]; EN 62479 [ed.]; EN IEC 63000 [ed.]
   Frequency band 2400-2483.5 MHz, max radio-frequency power <value from test report> dBm.
7. Notified body: none (Annex II internal production control)  [if used: name, no., certificate]
8. Additional information: Accessories and components needed for compliance: power cord (3 m), ...
   Software version covered: <fw x.y>. 
Signed for and on behalf of: <name, function>, Place/date: <...>

1.4 Technical-file outline (RED Annex V / GPSR / RoHS)

1. Product description, variants, photos, block diagram, intended use, user groups (professional and prosumer).
2. Design drawings, PCB layouts, schematics (mains side highlighted), BOM with safety-critical parts (heater, thermal cut-out, relay, fuse, Y-caps, inlet, cord, 24 V module with approvals).
3. Safety-critical function description: hardware limiter, relay logic, firmware limit, fault tests (links to design-basis).
4. Risk assessment (EN ISO 12100-style hazard list incl. burns, fire, aerosol/slip, misuse) - GPSR Art. 9 / ZAD-SAF requirements.
5. Standards list with editions applied and which clauses; test reports (safety, EMC, radio, EMF), module reports and FCC/RED evidence for the module.
6. Radio: module data, antenna type/gain, firmware controls limiting power, frequency channels, coexistence/adaptivity statements; Art. 3.3 assessment note (see 0.5).
7. RoHS: material declarations, exemptions, EN IEC 63000 assessment; REACH SVHC survey (date of candidate list used).
8. Labels and markings drawings, manual (languages), packaging; ZAD-SAF-08 safety text (section 4 below).
9. Production control: inspection plan (100 % hipot/earth-bond/ functional test in EMS, ASSUMPTION), serial/batch rules, change control for firmware and the BOM, record of non-conformity.
10. DoC(s), copy of CE label, importer/distributor information, retention 10 years.

2. US and Canada

2.1 Findings of the quick web check on the NRTL standard

• SOURCE (https://webstore.ansi.org/standards/esta/ansie1292009r2014 and the ESTA page, https://tsp.esta.org/tsp/documents/published_docs.php, accessed 2026-10-30): ANSI E1.29 "Product Safety Standard for Theatrical Fog Generators..." is based on ANSI/UL 998-2006 Humidifiers with modifications, and is intended to guide testing laboratories such as UL and ETL; the edition shown by the second source is ANSI E1.29-2009 (R2024) (older R2014/R2018 superseded). One search result also lists a commercial fog machine with "UL 998 / CSA C22.2 No. 104(?)" on its spec sheet: the Canadian number was truncated in the result ("CSA C.22.2 No."), so C22.2 No. 104 is NOT confirmed by this check (see Martin spec URL above; edition of UL 998 to confirm).
• SOURCE (https://metlabs.com/about/standards/nrtl-standards/, accessed 2026-10-30): UL 998 Humidifiers is on the list of standards NRTLs (e.g. Eurofins MET) can certify to, so UL, ETL (Intertek), MET or CSA are all possible. Conclusion: the requirement ZAD-CRT-02 statement (UL 998 with E1.29) is consistent with what I found; the CSA C22.2 No. 104 number and the current UL 998 edition remain "to confirm" with the lab.
• UL 8750 is not relevant to the fog generator itself (only to LED parts, author's knowledge).

2.2 FCC and ISED plan

2.3 NRTL safety listing (UL / ETL / CSA)

2.4 California Proposition 65

Safe Drinking Water and Toxic Enforcement Act 1986: a warning is required if the product exposes California consumers to listed chemicals above the safe harbour level; the usual triggers here are lead in brass/solder, DEHP/DINP and other phthalates in cords, BPA. Measure: BOM review, low-lead cords (RoHS-aligned supplier declarations give a good base), then the on-box/web-shop warning ("WARNING: Cancer and Reproductive Harm - www.P65Warnings.ca.gov") if exposure cannot be excluded. Cost 500-2,000 for the review (ASSUMPTION); no filing is made with the State. List of chemicals updates yearly: re-check at the time of shipping.

3. ESTA / ANSI entertainment-industry standards (ZAD-CRT-03)

Editions SOURCE: https://tsp.esta.org/tsp/documents/published_docs.php and the search summary of 2026-10-30 (https://blog.ansi.org/ansi/ansi-e1-23-entertainment-technology-fog-effects/); where sources disagree "edition to confirm".

4. Information duties of ZAD-SAF-08 (labels and manual)

Manual and labels (EN 60335-1 cl. 7 and ANSI E1.14, ZAD-SAF-08) carry at least:

1. Smoke detectors and fire alarms: the fog can trigger them; inform the building/ fire authority, isolate detectors only per local rule.
2. Ventilation and exposure guidance (E1.5 / E1.23); do not use in small closed rooms; visibility.
3. Slippery residue on floors; clean-up.
4. Minimum distance from the nozzle to persons (value from the test, TBD by the thermal test) and to flammable materials/ curtains.
5. Hot surfaces: warning symbol on nozzle and exchanger area; cool-down time statement (value from test).
6. Approved fluids only (E1.5-class water-glycol/glycerin); no oils, perfumes, additives; fluid tank handling.
7. Mains: Class I, earth required, voltage per variant, fuse type (EU T6.3 A H, US T10 A), do not open, qualified service, no operation with damaged cord.
8. Not for use with persons with asthma/ respiratory issue present without assessment; keep out of reach of children; indoors, dry (IP rating if declared).
9. Radio: frequency band and max power, FCC 15.19 / 15.105 text, ISED statements, "Contains FCC ID / IC", CE mark, WEEE and battery symbols, battery insertion polarity and disposal for the remote.
10. Supplier/manufacturer name and address, model, serial/batch, rating label (V, Hz, W), date of manufacture. Label drawings and the manual safety section go into the technical file (1.4 item 8).

5. Summary table (cost and calendar time; all ASSUMPTION unless stated)

Per unit at 1,000 units (CALCULATION): 33.2 to 74.1 EUR/unit, which is 55 to 123 % of the 60 EUR target unit cost: the certification budget must be funded separately from the BOM and be in the Kickstarter budget. Two variants (EU, US) share the platform but have separate heater/ fuse/ cord, so safety tests are per variant (the totals above assume the second variant needs delta tests only: ASSUMPTION).

Calendar time (ASSUMPTION): design freeze on pre-production samples -> pre-compliance 2-3 weeks -> EU lab tests 6-10 weeks (safety longest; EMC/radio in parallel) -> DoC/file 1-2 weeks: about 3-4 months for the EU. US: listing project 8-16 weeks, FCC/ISED 2-4 weeks in parallel: about 3-5 months. Both in parallel with shared samples: 4-6 months from frozen pre-production units, plus a 1-2 month reserve for failures. Certification must be completed before the first Kickstarter delivery and the sample units must be production-representative (tooling-free sheet-metal enclosure helps). Order: antenna/metal decision (0.1) -> prototype B -> pre-compliance -> formal tests.

Benchmark — comparable commercial fog machines, 2026-10-30

Method: web search / retailer pages, all accessed 2026-10-30. Time-boxed (about 8 look-ups), so many cells are "n/a". Nothing was invented. Legend: SRC = SOURCE (URL, accessed 2026-10-30). ASSUMPTION always carries a reason. "n/a" = value not read. The WebSearch tool returns summaries, not page text. Each SRC below is the retailer or manufacturer page named in the result, not a page I read in full. Fullcompass.com returned HTTP 403 when fetched, so its values come from search-result summaries only.

Sources

• S1 https://www.fullcompass.com/prod/157407-antari-z-1200-ii-1200w-water-based-fog-machine-with-dmx-control-18-000-cfm-output (via search summary)
• S2 https://djdeals.com/product/antari-z-1200-ii/ (price USD 517.95, via search summary)
• S3 https://www.chauvetdj.com/products/hurricane-1800-flex/ (official; via search summary)
• S4 https://www.fullcompass.com/prod/549109-chauvet-dj-hurricane-1800-flex-water-based-fog-machine-with-adjustable-angle-29-600-cfm-output-and-dmx (USD 399.99 and 29,600 cfm, via search summary)
• S5 https://www.idjnow.com/chauvet-dj-hurricane-1800-flex-dmx-controllable-fog-machine.html (second spec set, via search summary)
• S6 https://www.thomannmusic.com/cameo_phantom_f5.htm (fetched; the US-store price shown was USD 265)
• S7 https://www.thomannmusic.com/adj_fog_fury_jett.htm (fetched; the US-store price shown was USD 233)
• S8 https://www.adj.com/products/fog-fury-jett (official, via search summary)
• S9 https://images.thomann.de/pics/atg/atgdata/document/manual/c_355877_355990_355991_v6_r1_en_online.pdf (Stairville M-Fog 1000/1500/2500 manual, via search summary)

Table

Continuous capability implied by heater power

Formula: ml/min = P x 60 / e, with e = 2.4 kJ/ml (ASSUMPTION: water-glycol fog fluid, heated from ambient to vapour; the design basis uses 2.35 to 2.55 kJ/ml). So ml/min = P/40. This ignores losses and is therefore an upper bound.

Takeaways for our design

• Most claimed fluid rates are 2.4x to 7x above what the heater can vaporise continuously (Z-1200 II, Phantom F5, Jett; the Hurricane only in its 80 ml/min spec set). Those claims are peak figures, drawn from stored block heat. The Phantom F5 states it openly: about 14 s of fog, then about 12 s of post-heating.
• The Jett's 120 ml/min on 650-700 W is physically impossible as a continuous rate (about 16-17 ml/min at most). Treat "m3/min" and "cfm" claims as unverifiable marketing numbers. We should state output in ml/min of fluid plus a defined test, not a bare cfm.
• Our 950 W implies a ceiling of about 24 ml/min continuous. That is below Z-1200 II and Phantom F5 (30 and 37.5) but with a 0.55 kg low-mass block it is a different regime. The "pro output, no reheat pauses" requirement must therefore be met by derating smoothly to q_cont, as the design basis says, and by publishing that figure honestly. If the brief expects Phantom-F5-class output, 950 W is probably about 1.5x short. The heater power should be revisited in the estimates.
• Competitors at 130-230 EUR class (Phantom F5, Jett; about USD 233-265 at the Thomann US store, EUR not read) have warm-up of 4-9 min. A warm-up of 2-3 min (the Stairville manual says about 135 s at 1500 W) would be a differentiator. Our design needs roughly 0.55 kg x 0.9 kJ/(kg K) x 230 K, about 114 kJ for the block, which is about 2 min at 950 W (ASSUMPTION: aluminium cp 0.9 kJ/kg K, from 20 to 250 C, losses and coil fluid ignored).
• All benchmarks use wired timer remotes, or (Jett) a wireless one, with 3-pin DMX and 1 to 7 channels. None of the sources I read mention a 5-pin DMX connector. Two-way wireless status is open ground. Competitor DMX is simple (1-3 ch), so a small channel set is acceptable.
• Tank sizes are 1.8 to 3.8 L and masses are 6.75 to 10.9 kg. Our 1.3 L and 2.8 kg are lighter but the tank is smaller. At 24 ml/min continuous, 1.3 L lasts about 54 min; check against the brief's run-time requirement.
• Data gaps (n/a): EUR street prices with shop URLs, manual warm-up and reheat specs, review-based spitting evidence for Antari and Chauvet. EUR prices must be filled in before the price positioning against EUR 229 is relied on.

Design basis — concept C2 (chosen), rev A, 2026-10-30

Every value below is an ASSUMPTION or a CALCULATION at concept level; final numbers come from 02-concept/estimates.py → estimates.results.json. Requirement ids are those of INPUT/requirements.md.

Working principle (C2 — "metered flow-through exchanger with energy-balance control")

• Low-mass die-cast aluminium heat exchanger (≈0.55 kg AlSi alloy) with a cast-in tubular heater element and a cast-in stainless 316 fluid coil (≈1.2 m, 6×1 mm), insulated with 20 mm ceramic-fibre/microporous board in a steel shield box. Heated nozzle (3 mm bore) screwed into the block outlet, guarded.
• Heater 950 W per variant: EU 230 V (≈55.7 Ω), US 120 V (≈15.2 Ω). Rated input ≈960 W (EU and US).
• Fluid metered by a stepper-driven peristaltic pump (24 V DC bus, TMC2209-class silent driver): 1–60 ml/min, self-priming, dry-run safe, flow independent of mains voltage; metal spring check valve at the exchanger inlet; sintered inlet filter reachable without tools.
• Control: pump rate = min(command, energy budget). Continuous rate q_cont = (P_heater(V_measured) − P_loss) / e_fluid; stored heat above T_min may be spent as peak, then the rate derates smoothly to q_cont — never a pause (ZAD-PRF-03). No pumping below T_min (no spitting). Fluid-type setting (Quick / Standard / Dense) sets T_min / T_ready / T_set and e_fluid.
• Temperatures (Standard fluid): T_min 240 °C, T_ready 250 °C, T_set 290 °C, firmware limit 300 °C, independent hardware limiter 320 °C (second PT1000 + comparator → opens series mains relay, latching), non-self-resetting thermal cut-out ≈340 °C on the block (last resort, EN 60335-1 cl. 19).
• Energy per ml: typical water-glycol fluid 2.35 kJ/ml, worst (water-rich) 2.55 kJ/ml (ASSUMPTION).
• Heater switching: zero-cross triac (integral-cycle control, flicker-limited switching) + series safety relay driven only by the hardware limiter; MCU measures mains RMS voltage.
• Electronics: certified 24 V 15 W AC/DC module (reinforced insulation, mains ↔ SELV); main MCU + radio = pre-certified 2.4 GHz BLE module (ESP32-C3-MINI-1 class, FCC ID / RED report) with BLE Long-Range (coded PHY), two-way; DMX512-A/RDM via RS-485 transceiver behind digital isolator + isolated DC/DC; 5-pin and 3-pin XLR in + thru; 1.3" OLED + encoder + 2 keys; RGB status LED visible from 5 m; USB-C service port (firmware update, SELV); capacitive low-fluid sensor on the tank + optical bubble sensor on the suction line.
• Remote: handheld, same 2.4 GHz module, 2×AAA, unique pairing ID, 4 machines / groups, two-way status LED.
• Tank 1.3 L usable, translucent PP/HDPE, wide 38 mm cap, ePTFE vent membrane (no leak in any orientation), weighted flexible pick-up (tilt ±30°), backlit level window.
• Enclosure: powder-coated aluminium sheet chassis + cover (no mould tooling), Class I, all metal + exchanger bonded to PE; vent slots; no fan. Bounding box ≈300 × 170 × 200 mm (10.2 L); dry mass ≈2.8 kg.
• Mains: IEC C14 inlet with 2-pole switch + fuse (EU T6.3 A H, US T10 A), 3 m detachable cord (EU Schuko–C13 H05VV-F 3G1.0; US NEMA 5-15P–C13 SJT 18 AWG); Y-capacitors ≤1 nF.
• Production: 1,000 units, EU assembly (Poland EMS), sheet-metal enclosure; target unit cost ≤60 EUR (estimate ≈75 EUR ±30 %), retail 229 EUR incl. VAT.

Rejected concepts

• C1 classic high-mass block + AC vibratory pump, on/off thermostat (budget-machine architecture).
• C3 induction-heated flow-through tube (no block).
• C4 twin exchangers firing alternately.

Hazard analysis, safety requirements, concept DFMEA and fault-test list — concept C2, rev A, 2026-10-30

Basis: 02-concept/design-basis.md (C2 metered flow-through exchanger), INPUT/requirements.md ZAD-SAF-01..08, ZAD-PRF-07, ZAD-FUN-06, ZAD-REL-01. Labels: SOURCE = taken from a cited standard/table; CALCULATION = derived here; ASSUMPTION = engineering judgement to be verified in the prototype. Scales: Severity S 1 negligible, 2 minor injury/first aid, 3 serious injury (burn needing treatment, shock), 4 fatal/fire/multiple injuries. Probability P (hazard analysis) 1 remote, 2 unlikely, 3 possible, 4 likely. Risk = S × P (1–4 low, 6–8 medium, 9–16 high; high is not acceptable, medium needs documented ALARP). Method: ISO 12100 three-step reduction: (1) inherently safe design, (2) safeguarding / protective measures, (3) information for use. Standards named here (EN 60335-1, EN 60335-2-?? for the applicable part, IEC 61032, ANSI E1.5, E1.14, E1.23) are the ones the requirements cite; edition and the exact particular standard are to be confirmed by the compliance engineer (ASSUMPTION).

1. Hazard analysis

Calculation — pressure in the coil with a blocked nozzle (H-09)

• SOURCE: IAPWS-IF97 saturation line of water: p_sat(290 °C) ≈ 7.44 MPa ≈ 74 bar; p_sat(320 °C) ≈ 11.3 MPa ≈ 113 bar; p_sat(340 °C) ≈ 14.6 MPa ≈ 146 bar (values read from IAPWS-IF97 steam tables, to be re-checked in estimates.py).
• CALCULATION: SREQ-04 rating is ≥100 bar at 350 °C. A blocked outlet with a water-rich fluid at the T_set of 290 °C gives about 74 bar from vapour pressure alone, which is below the rating (margin ≈ 1.35). At the limiter trip 320 °C the vapour pressure is about 113 bar, which is above 100 bar; at the 340 °C cut-out it is about 146 bar. If the coil is completely liquid-filled (no vapour space), pressure is set by the pump stall force and thermal expansion, not by Psat, and can be higher still.
• Consequence for design (recommendation): rate the hot-side path ≥150 bar at 350 °C, or add a pressure relief to a safe, contained discharge; blocked-outlet detection must stop the pump long before 74 bar. The fluid is a water–glycol mix, so the vapour pressure at a given T is lower than for pure water (Raoult, ASSUMPTION); the pure-water figure is used as the worst case.
• ASSUMPTION: 316 tube 6×1 mm, Barlow estimate P = 2·S·t/D with S ≈ 115 MPa at 350 °C (ASSUMPTION, check in ASME II-D) gives ≈ 380 bar burst-order; the weak points are the fittings, the check valve and the cast-in transition, not the tube. To be proven by a hydrostatic proof at 1.5 × rating and a burst test on the prototype.

2. Safety requirements derived from the table

Targets for lifetime (ZAD-REL-01): safety functions (limiter, relay, cut-out) are verified over the rated heater/pump life; the relay is rated for the number of limiter trips in a 1,000-unit field plan (ASSUMPTION), and the cast-in heater and pump tube have a replacement interval written into the manual.

3. Concept DFMEA

Ratings: S severity 1–10, O occurrence 1–10, D detection 1–10 (10 = not detectable); RPN = S×O×D. These are concept-level ASSUMPTIONS; action is required for RPN ≥ 100 or S ≥ 9 regardless of RPN.

3.1 Heater / exchanger

3.2 Pump / fluid path

3.3 Control electronics & firmware

Coverage check for ZAD-SAF-01 single faults: sensor open (3.1 row 1), sensor short (3.1 row 2), heater switch stuck on (3.1 row 4, 5), firmware hung (3.3 row 1), pump failure (3.2 row 1), empty tank (3.2 row 4), blocked outlet (3.2 row 6). Each has a measure that works without relying on the failed element.

4. Fault-test list for the prototype test plan (EN 60335-1 clause 19 abnormal operation)

General: test at 1.06 × rated voltage (EU 243.8 V; US 127.2 V) (CALCULATION from 230 V/120 V), enclosure on a wooden test corner as required by clause 19 (ASSUMPTION), thermocouples on block, shield, accessible surfaces, nozzle area, wood board; each fault is applied one at a time. Common pass criteria (ZAD-SAF-01): no flames, no molten material ejected, no emission of hot liquid, enclosure temperatures within the clause 19 abnormal-operation limits, after the test the insulation withstands the dielectric strength test and the leakage current is within limit; PE bonding still ≤0.1 Ω. Parts and cut-outs that opened are replaced between tests as the standard allows.

5. Residual risks the design cannot remove

1. A hot aerosol jet leaves the nozzle by function. Anyone who puts a hand, face or eyes into the near field can be scalded or irritated; the design reduces it (flow limit, guard, minimum distance) but cannot remove it. Information for use only.
2. Fog triggers smoke detectors and fire alarms, reduces visibility on exit routes and leaves a slippery film; no technical measure inside the machine prevents it. The venue operator must manage detectors, exits and floor cleaning (ANSI E1.14, SREQ-12).
3. Fluid exposure (inhalation, skin, eyes) depends on the fluid, dose and room ventilation, which the machine does not control; fluid composition is the user's responsibility (E1.5, E1.23).
4. Hold mode and continuous mode are chosen by the user; an unattended machine in these modes is a user decision and is only reduced by the auto-off for the remote mode.
5. Hot surfaces: the nozzle guard and its surroundings stay hot (excepted in ZAD-SAF-02); the exchanger remains hot for some minutes after switch-off and refill while hot is possible when the warning is ignored.
6. Blocked-outlet pressure: at 320 °C, p_sat of water is about 113 bar, above the 100 bar rating; this is closed by SREQ-14 (rating ≥150 bar or relief). If that is not adopted, a residual rupture risk remains in a double fault (blocked outlet + failed detection).
7. Common-cause failures are not covered by the single-fault concept (fire in the building, water poured in, deliberate bypass of the limiter or cut-out, use of unapproved fluids).
8. Radio coexistence in a 2.4 GHz-saturated venue can delay or lose the link; safe state is output stop, but unwanted outputs from a stuck remote cannot be excluded beyond auto-off.
9. Cell battery ingestion by children is reduced by a screw-secured cover but not eliminated.

All numbers above are concept-level; update after estimates.py and the prototype tests, and review again with the compliance engineer (notified body or test house) before design freeze.