PULSCAR ANALYSIS - P0017, P0012 AND P0018 (CRANKSHAFT TO CAMSHAFT CORRELATION, CAM TIMING, CAM PHASER AND TIMING CHAIN) IN NHTSA'S PUBLIC FILES PulsCar analysis of NHTSA public data, October 2026. License: CC BY 4.0 Attribution: "PulsCar analysis of NHTSA public data, October 2026." Published alongside https://www.pulscar.io/blog/p0017-code ------------------------------------------------------------------------------ Prepared October 2026. Source data: NHTSA public flat files downloaded on 7 October 2026 (complaints received 1 January 2015 to 4 October 2026 - the latest received date in the files - 782,908 records; manufacturer communications; recalls since 2010; investigations) and documents downloaded one at a time from static.nhtsa.gov. 1. Complaints (6,766, each sorted twice, independently) First search: every complaint whose narrative names a camshaft-timing family code by any spelling we searched (P0011 P0012 P0013 P0014 P0015 P0016 P0017 P0018 P0019 P0021 P0022 P0024 P0025; a letter O for zero, a space, hyphen or period after the P, the suffix 00, or a bare number after "code", "DTC", "error", "CEL" or "fault"), names the timing chain (timing chain, chain stretch, chain tensioner, chain guide) together with a light, code, rattle, jump, skip, phaser or timing context, names a cam phaser or variable valve timing part (phaser, VVT, VVT-i, VCT, cam actuator or adjuster, oil control valve, VANOS, valve timing) together with a light, code, rattle, knock, stall, misfire, rough-running, limp or chain context, or contains the word "correlation" with a crank, cam, timing or code context: 3,549 complaints. The exact search patterns are in the sorting rules (the sorting rules file in this folder). Completeness check: before the second batch was sorted, an independent search of the whole complaint file for what the first search had missed proposed 3,217 more (about 2,000 name the timing chain without a context word; about 480 name a cam phaser or VANOS unit without one; about 260 use words the first search did not list, such as Honda "VTC actuator" and Hyundai or Kia "CVVT"; the rest use other wording such as "jumped time"). All 3,217 were added; their relevance was decided by the unchanged sorting rules and an addendum written before that batch was sorted (see the sorting rules file). The complaints file marks them found_by = completeness check. Sorting: the sorting rules (the sorting rules file in this folder) were written, dated and fingerprinted before any complaint was sorted. Two independent sortings covered all 6,766 complaints, each from the written rules only, without seeing the other. One category per complaint (N not a timing-chain, cam-phaser or cam-timing event on this car; T timing chain, tensioner, guides or sprockets; P cam phaser, VVT or VANOS unit; S oil control valve or VVT solenoid; K crank or cam position sensor or its ring; L oil; W software or a campaign procedure only; E engine failed, replaced or rebuilt with no single part named; O another cause; C checked and cleared; U nothing diagnosed or only the owner's guess) and five yes/no flags (damage, stall, noise, recall, returned). The two sortings matched on: category 6,063 of 6,766 (89.6%); damage 6,663 (98.5%); stall 6,623 (97.9%); noise 6,710 (99.2%); recall 6,600 (97.5%); returned 6,624 (97.9%). Relevance (N or not N): 6,707 of 6,766 matched. 1,178 complaints differed in the category or at least one flag; every one was read in full and settled under the written rules R1 to R11 and R2b (first batch) and C1 to C11 (second batch), published in the sorting rules file. A final check found 158 complaints on which both sortings agreed but the cause had come from the owner alone, with no shop, dealer or mechanic named and no part replaced; their category was changed to U. The resolution and rule_applied columns record all of it. Final: not relevant (N) 336; of the 6,430 relevant: T 2,499, U 1,843, P 1,104, E 379, S 192, O 160, L 127, C 55, W 38, K 33. First search 3,331 relevant, completeness check 3,099 relevant. Flags among the relevant: damage 1,548, stall 2,356, noise 1,942, recall 2,063, returned 972. 610 of the relevant name a family code themselves: P0017 179, P0016 170, P0014 156, P0011 111, P0012 42, P0018 39, P0024 27, P0021 24, P0019 20, P0013 19, P0022 14, P0015 7, P0025 2 (a complaint can name more than one). Makes among the 6,430: Ford 1,313, Chevrolet 1,064, BMW 540, GMC 441, Honda 389, Volkswagen 387, Hyundai 284, Kia 270, Nissan 255, MINI 169. Models: F-150 914, Equinox 483, Traverse 242, Terrain 241, Expedition 216, CR-V 214, 328i 182. Median model year 2013. 3,929 complaints give a mileage from 1,000 to 400,000; median 89,000 (timing chain set 91,000, cam phaser 80,000). By year received: 2015 434, 2016 573, 2017 570, 2018 607, 2019 686, 2020 512, 2021 317, 2022 384, 2023 536, 2024 696, 2025 697, 2026 418 (2026 to 4 October). In NHTSA's own fields: crash 25, fire 25, injury 18 complaints (28 people), towed 334. One record carries 40 deaths and 38 injuries in those fields, but its narrative is only a video link and a few words about a cam phaser; it is left out of the counts above. Mention counts (a search of the narratives for words, not sorted; case-insensitive, in the relevant narratives): phaser 988; VANOS 225; VTC 317; oil consumption or an engine burning oil 369; "rattle" 932; tensioner 608; "class action", lawsuit or settlement 314; "extended warranty", "warranty extension" or "extension" 235; refusal, denial or "not covered" wording 566; "timing belt" 117 (mostly the wrong word for a chain). Recall numbers named in the relevant narratives: 23V-707 64, 14V-176 29, 13V-395 11, 07V-519 5, 17V-368 4. Figures count complaints, not vehicles. 2. Owner-reported amounts (1,403 figures in 1,191 complaints) Every dollar figure (and "dollars" or "bucks" amounts, and a few "NNk" amounts) in the relevant complaints was pulled out by a text search (2,124 candidates: 1,215 in the first batch and 909 in the second) and decided one by one under one written rule: keep every figure the narrative ties to the diagnosis or repair of the timing fault - a diagnosis, inspection or teardown fee, a timing chain, tensioner, guides or kit, a cam phaser, VVT unit or camshaft, an oil control valve or solenoid, a position sensor, an oil service, an engine, long block, short block, rebuild or head - whether paid, billed or quoted; exclude mileages (81), vehicle prices, payments, trade-in and loan figures, rentals and lost wages (133), repairs of other systems (196), restatements of a figure already counted (167), other owners' costs and class-action sums (72), warranty limits, deductibles and refund demands (42) and figures whose subject cannot be told (30). 870 figures are quotes, 250 bills, 171 described as paid, 81 not stated and 31 offers or shares. Groups: chain 443 (median $2,500); engine 373 (median $8,000); several parts together 261 (median $4,000); cam phaser, VVT unit or camshaft 158 (median $3,514.20); diagnosis, inspection or teardown 86 (median $286.25); other or not itemised 46 (median $2,900); oil control valve or solenoid 20 (median $582.50); position sensor 10 (median $320); oil work 6 (median $350). Counting only bills and amounts paid: chain 134 figures (median $2,290.92), cam phaser 43 (median $3,400), engine 70 (median $6,709.22). 3. Manufacturer documents (708 files, 685 distinct) First search: every manufacturer communication in NHTSA's index (5.3 million index rows, 188,586 distinct communications) whose summary names a camshaft-timing family code (including the six-character form such as P001700), the timing chain with a symptom or part wording, a cam phaser, variable valve timing, oil control valve or cam adjuster wording, or the words correlation, crankshaft position or camshaft position: 664 NHTSA ids. Of these, 155 were set aside without being opened (43 motorcycles, powersports or heavy trucks; 112 whose summary matched only on a passing word, such as a general software-update or an unrelated repair), 78 are near-identical wordings of ids that were downloaded (grouped by 5-word overlap of 0.6 or more), and 431 were downloaded. Completeness check: an independent search of the same index proposed 245 further ids; 199 were downloaded and 46 are near-identical wordings of documents already downloaded (the document list marks 202 files as found by this check). 10 first-search ids had no PDF on NHTSA's site. Recalls and investigations: the filings of seven recall campaigns on the parts in this guide (BMW 14V-176 and 23V-707, Toyota and Lexus 13V-395, Porsche 17V-368, Chrysler 16V-907 and 18V-049, and Ford 17V-672, which proved not relevant): 77 recall documents (Part 573 reports, owner letters, interim letters, dealer notices, model breakouts, quarterly completion reports, NHTSA acknowledgments, chronologies); other recall papers in those folders were not read. Investigation records read from NHTSA's investigation file (no PDF): EA06-009, PE24007, DP17003, RQ22002, EA25004 and PE22005 (the last two proved not relevant). 708 files in all: 625 manufacturer documents, 77 recall documents and 6 investigation records. 23 are exact copies of another file, so 685 distinct. 506 were read in full; 189 are revisions, twins or copies that were compared line by line with a document read in full (every line that differed was read); 13 recall letters that are scanned pages were read as page images. Documents whose pages are pictures were read from page images too (a few Lexus, Audi, BMW and Mazda files). 78 files print P0017, 22 print P0012 and 47 print P0018 in their own text, counting six-character forms and revisions as separate files (files whose pages are pictures cannot be searched this way); no recall file prints P0017, P0012 or P0018 (the Chrysler 16V-907 chronology names P0335 and P0339). Every quotation in the article was checked word for word against the document or complaint text; quotes from pages that are pictures were read from the images and could not be checked that way. Coverage status in October 2026 is our computation from the terms printed in each document, not a statement by the manufacturer. No document prints an individual car's in-service date. Limits: complaints over-represent failures, refusals and expensive repairs; owners report diagnoses as they understood them; the Chevrolet, GMC, Ford and BMW complaint counts are large because those engines are common and long-lived, not because they fail more; the complaints file includes many for engines (for example Kia and Hyundai engines condemned after oil loss) where the timing code is a symptom of a larger failure; a narrative that cites only a campaign number is not counted as a recall mention unless the campaign is named; searches of NHTSA's index cannot see documents whose summary names none of the search terms; the 155 set aside and the 78 and 46 near-identical wordings that were not downloaded were not opened. Nothing here is a failure rate.