PulsCar analysis of NHTSA public data, September 2026 The cooling fan: 1,307 owner complaints, 13 recall campaigns covering 857,119 vehicles, and 406 manufacturer communications Licensed CC BY 4.0. Queries run 18 September 2026. -------------------------------------------------------------------------- SOURCE FILES -------------------------------------------------------------------------- NHTSA ODI complaint flat files COMPLAINTS_RECEIVED_2015-2019, _2020-2024, _2025-2026 -- https://static.nhtsa.gov/odi/ffdd/cmpl/ (dictionary CMPL.txt) 777,040 distinct vehicle complaints (ODINO, product type V) received 1 January 2015 to 8 September 2026. NHTSA recall flat file FLAT_RCL.zip (one row per make/model/model year per campaign; counted here as distinct campaign numbers). Every campaign quoted in the article was also re-read live at https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=... NHTSA manufacturer communications flat file, 5,267,097 rows. -------------------------------------------------------------------------- WHICH COMPLAINTS COUNT (the cooling-fan set: 1,307) -------------------------------------------------------------------------- The narrative contains one of these exact phrases, case-insensitively: COOLING FAN | RADIATOR FAN | ENGINE FAN | FAN MOTOR | FAN CLUTCH | FAN RELAY | FAN CONTROL MODULE | FAN RESISTOR Motorcycle, heavy-truck, bus, RV, fire-apparatus and ambulance makers are excluded by make. Result: 1,307 complaints. WHY THE QUERY IS PHRASE-BASED AND NOT JUST "FAN": the word fan on its own matches blower fans, condenser fans, seat fans, radio head-unit fans and laptop-style cooling fans in infotainment units. Requiring a two-word engine-cooling phrase costs some recall and buys most of the precision. SAMPLE READING (done before any count below was published) Thirty matches drawn at random (seed 31) and read in full. Twenty-nine described an engine cooling fan. One did not (a Honda Accord narrative about a cooling fan inside the radio head unit). FALSE-POSITIVE RATE OBSERVED: 1 IN 30, ROUGHLY 3%. A SET WE TRIED AND DID NOT PUBLISH We also built an "overheats at idle" set: an overheating word within 120 characters of an idle, stoplight, traffic or stop-and-go phrase, with engine-cooling context required and non-engine overheating (brake-light connectors, transmission clutches, control modules, starters) excluded. It returned 195 complaints, and reading a random thirty showed that most were general overheating narratives that mention traffic in passing rather than the idle-only pattern this page is about. The tighter version -- a narrative that ALSO says the car cools down once it moves -- returned THREE. Owners simply do not describe the pattern in those words often enough to measure. Neither set is quoted in the article as a count; the article says plainly that the signature is rare in the file's language. -------------------------------------------------------------------------- RESULT 1 -- WHAT OWNERS NAME (fan-complaints-what-owners-name.csv) -------------------------------------------------------------------------- Shares of the 1,307. A narrative can name more than one thing, and the share measures vocabulary, not cause frequency. the fan motor or the whole fan assembly 228 17.44% wiring, a connector or a harness 72 5.51% a fuse 65 4.97% a fan control module or resistor 52 3.98% a fan clutch 33 2.52% a fan relay 31 2.37% a temperature switch or coolant temperature sensor 18 1.38% a fan blade that broke or came apart 5 0.38% Behaviour described: the fan did not come on / was inoperative 80 6.12% the fan ran constantly or at full speed 17 1.30% overheating is mentioned 340 26.01% the air conditioning is mentioned 256 19.59% smoke, melting, a burning smell or fire near the fan 130 9.95% Outcome flags, cooling-fan set vs all 777,040 complaints: FIRE 5.05% vs 1.82% (66 complaints carry the fire flag) towed 4.06% vs 5.11% crash 0.92% vs 4.37% median miles 58,000 (745 complaints report a usable figure, 0 < miles < 400,000) vs 70,000 across the file. THE FINDING AND ITS LIMIT: a cooling-fan complaint is about 2.8 times as likely to carry the fire flag as the average complaint in the file, and LESS likely to involve a tow or a crash. The limit is selection: a fan that merely stops working is an inconvenience people take to a shop, while a fan circuit that melts is the kind of event people report to the federal government. The recall record below is the independent check on that, and it points the same way. -------------------------------------------------------------------------- RESULT 2 -- RECALL CAMPAIGNS (fan-recall-campaigns.csv) -------------------------------------------------------------------------- Every campaign whose defect, consequence or component text names a cooling fan, radiator fan, engine fan, fan motor, fan clutch, fan relay, fan control module, fan resistor, fan blade or fan module, after removing heavy-truck, bus, fire-apparatus, ambulance and motorcycle manufacturers. 13 distinct campaigns. 857,119 vehicles by the manufacturers' own potentially-affected counts, summed once per campaign. 10 of the 13 (77%) name FIRE in the defect or consequence text. Largest: 25V-720 (Ram ProMaster, 300,691), 19V-818 (Ram ProMaster, 99,128), 15V-015 (Kia Forte, 86,880), 16V-868 (Chevrolet Tahoe police vehicles, 49,927), 12E-029 (Valeo aftermarket replacement fans for BMW, 585). -------------------------------------------------------------------------- RESULT 3 -- MANUFACTURER COMMUNICATIONS -------------------------------------------------------------------------- FTS5 over bulletin summaries for the same phrase list, then the phrase re-checked in the summary text, then commercial and motorcycle brands removed: 406 distinct communications. 83 of them (20.4%) name a software update or reprogramming; 66 (16.3%) describe a noise, rattle, roar, whine, hum or vibration. Summaries are abstracts, so both shares are floors. -------------------------------------------------------------------------- PRIMARY DOCUMENTS READ, 18 SEPTEMBER 2026 -------------------------------------------------------------------------- Recall 25V-720, Chrysler (FCA US, LLC), reported 23 October 2025 -- 300,691 2018-2026 Ram ProMaster; "The electrical circuit for the radiator fan may overheat"; "Electrical circuits that overheat increase the risk of a fire"; remedy replaces the cooling fan module and fuse. https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=25V720000 Recall 19V-818, Chrysler, 15 November 2019 -- 99,128 2015-2018 Ram ProMaster; "The engine cooling fan may seize which can cause the cooling fan motor to overheat"; "An overheated fan motor increases the risk of a fire"; dealers replace the engine cooling fan module free of charge. https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=19V818000 Recall 15V-015, Kia Motors America, 16 January 2015 -- 86,880 2014 Forte; "the cooling fan resistor may overheat and melt"; remedy replaces the cooling fan resistor and multi-fuse unit, with an ECU software update on 1.8L cars. https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=15V015000 Recall 12E-029, Valeo, 10 August 2012 -- 585 aftermarket replacement engine cooling fans (part numbers 696121, 696273, 696275) for 2000-2011 BMW models; a metal-oxide semiconductor may overheat, which "could cause the fan motor to become inoperative, or cause the motor circuit board in the assembly to overheat, potentially leading to a fire in the engine compartment"; "This recall is being conducted by Valeo, not by BMW." https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=12E029000 All four were re-read live from the NHTSA recalls API on 18 September 2026 and each returned HTTP 200. -------------------------------------------------------------------------- LIMITATIONS -- READ BEFORE QUOTING ANY NUMBER -------------------------------------------------------------------------- 1. Complaints are self-reported, unverified and voluntary. They are not failure rates; NHTSA publishes no denominator. 2. Counts favour high-volume models and models whose problem has been publicized, so the make and model table ranks attention as much as failure. Nothing here ranks reliability. 3. Keyword shares measure vocabulary. An owner who did not write "relay" may still have had one replaced. 4. The fire share is inflated by the same selection that inflates every frightening outcome in this file. The comparison that survives it is the one used: the fire flag in the fan set against the SAME field across all 777,040 complaints. 5. Thirteen recall campaigns is a small set and is reported as one. -------------------------------------------------------------------------- FILES -------------------------------------------------------------------------- fan-complaints-what-owners-name.csv row, complaints, share, median miles, miles reported, fire %, towed %, plus the set total and the file baseline. fan-complaints-by-make-and-model.csv makes and models with counts, shares, median miles and fire share. fan-recall-campaigns.csv campaign, date, manufacturer, makes, potentially affected, component, whether fire is named, defect text. No personal data is published: no VIN, no city, no dealer name, no owner free text beyond short quoted snippets in the article itself.