PULSCAR ANALYSIS — P0016 IN NHTSA OWNER COMPLAINTS: THE TIMING CHAIN SHARE, THE ENGINE-REPLACEMENT SHARE, AND THE 2025-2026 SURGE Pulscar analysis of NHTSA public data, September 2026. License: CC BY 4.0. Attribution: "Pulscar analysis of NHTSA public data, September 2026." Published alongside https://www.pulscar.io/blog/p0016-code ------------------------------------------------------------------------------ 0. WHAT THIS FILE ADDS, AND WHAT IT DELIBERATELY DOES NOT REPEAT ------------------------------------------------------------------------------ The Pulscar Check Engine Code Index already counts the MANUFACTURER SIDE of this code and publishes its own method and data: * 49 distinct manufacturer bulletins from 7 manufacturer groups name P0016; at least 13 of them (27 percent) are software, and 20 percent once the Stellantis flash bulletins are set aside. Top groups GM 21, Ford 11, JLR 5, Stellantis 5. Zero of the 49 is a warranty extension or emissions campaign. The index also documents GM's service updates N252510320 and N252510321, which add a Driver Information Center warning message when P0016 sets. https://www.pulscar.io/blog/check-engine-code-index https://www.pulscar.io/research/check-engine-code-index/code_index.csv Those figures are NOT recomputed here. This file answers the complaint-side questions the index does not: what owners name as the part, how often the complaint ends with an engine being replaced, how the set behaves on the safety flags, and why 43.8 percent of all P0016 complaints in eleven years arrived in the last twenty months. Where the index reports a median of 88,000 miles over 71 readings for P0016 and this file reports 87,730 over 72, the two pages are applying different odometer exclusions to the same 162 complaints; this file keeps every reading strictly between 0 and 400,000 miles and averages the two middle values of an even-numbered set. The difference is 270 miles and one reading. ------------------------------------------------------------------------------ 1. SOURCE FILES AND DATE RANGE ------------------------------------------------------------------------------ COMPLAINTS NHTSA ODI complaint flat files, product type V (vehicle) https://static.nhtsa.gov/odi/ffdd/cmpl/CMPL.txt (field layout) 777,040 distinct vehicle complaints, de-duplicated by ODI report number, received 1 January 2015 to 8 September 2026. TSBs NHTSA manufacturer communications flat files https://static.nhtsa.gov/odi/ffdd/tsbs/TSBS.txt (field layout) RECALLS NHTSA recall flat files, plus the live campaign API https://api.nhtsa.gov/recalls/campaignNumber INVESTIGATIONS NHTSA ODI investigation flat files (closing resumes) Data pulled 14 September 2026; this analysis run 19 September 2026. No third-party data was used. ------------------------------------------------------------------------------ 2. THE COMPLAINT SET ------------------------------------------------------------------------------ Query: the narrative field (DESCR) matched CASE-INSENSITIVELY against the whole-word regular expression \bP0016\b. Case-insensitivity is not optional: NHTSA narratives are stored ALL CAPS through 2020 and mixed case afterwards, and a capitals-only pattern silently drops the newer half of the file -- which for this code would be catastrophic, because 43.8 percent of the set arrived in 2025 and 2026. SELECT * FROM complaints WHERE descr LIKE '%P0016%'; -- 170 rows then kept only rows matching \bP0016\b -- 162 rows The eight rejected rows were read in full. Seven are run-together or extended strings in which the five characters sit inside a longer code: "P00166", "P001600" (four times), "P001685" and a list reading "P001477, P001600, P001762". The eighth is "DTCP0016" run together with the preceding letters. FIVE of these eight are almost certainly reporting this code -- "P001600" is the six-digit extended form GM and several scan tools print, and "DTCP0016" is a typing slip. They are excluded anyway, to keep this set identical to the one behind the Check Engine Code Index, which uses the same whole-word rule. The effect of including them would be 167 instead of 162, a 3.1 percent difference that changes no finding on the page. This is stated rather than quietly resolved. FINAL SET: 162 complaints. This is the denominator for every percentage below unless another denominator is named. Complaints are self-reported and unaudited. They are not failure rates, and there is no denominator in this data -- no sales, registration or fleet figure -- so no model here can be called better or worse than another. ------------------------------------------------------------------------------ 3. SAFETY FLAGS AGAINST THE WHOLE FILE ------------------------------------------------------------------------------ P0016 set (n=162) all 777,040 complaints crash flag Y 1 0.62% 4.37% fire flag Y 1 0.62% 1.82% towed flag Y 4 2.47% 5.11% injuries 0 deaths 0 Every flag is below the baseline. That is NOT the same as "this code is safe", and the page says so: 64 of the 162 narratives (39.5 percent) describe a stall, a shutdown or a loss of motive power, and 38 (23.5 percent) place the event on a highway, freeway or interstate. THE TWO INSTRUMENTS DISAGREE, LOUDLY. The structured towed field says 4 of 162 (2.47 percent). The word "tow" appears in 41 narratives (25.31 percent) -- ten times as many. Both numbers are published and neither is averaged. The structured field is filled in by the person completing the form and is plainly under-reported in this set; the narrative count is a floor of its own. ------------------------------------------------------------------------------ 4. ODOMETER ------------------------------------------------------------------------------ Readings kept only where 0 < miles < 400,000: 72 of the 162 reports state one. n = 72 minimum 73 lower quartile 72,900 MEDIAN 87,730.5 (middle values 86,000 and 89,461) upper quartile 113,750 maximum 297,000 at or under 36,000 miles: 3 of 72 ( 4.2%) at or under 60,000 miles: 8 of 72 (11.1%) at or under 100,000 miles: 45 of 72 (62.5%) at or under 120,000 miles: 59 of 72 (81.9%) The all-complaints median over the same filter is 70,000 miles (392,935 readings). The P0011 set on the sibling page sits at 76,222 over 60 readings, and only 8 of 72 P0016 reports fall inside a 5-year/60,000-mile powertrain warranty against 21 of 60 for P0011. Published as p0016-outcomes-and-mileage.csv. ------------------------------------------------------------------------------ 5. WHAT OWNERS NAME, AND HOW THE COUNTS WERE CHECKED ------------------------------------------------------------------------------ Each category is a regular expression run case-insensitively over the narrative. A complaint can fall into several categories; the columns do not sum to 162. These measure WHAT OWNERS WROTE DOWN, not what a wrench eventually found. timing chain, guide or tensioner 49 30.25% an engine replacement or rebuild (performed, quoted or required) 37 22.84% camshaft or crankshaft position sensor 20 12.35% engine oil: service, viscosity, sludge or filter 16 9.88% engine oil: level, consumption or overfill 15 9.26% engine oil pressure 12 7.41% cam phaser, actuator, sprocket or reluctor 9 5.56% VVT hardware (solenoid, oil control valve, VVT/CVVT/VTC) 5 3.09% THE HEADLINE CONTRAST. The identical patterns run against the 107-complaint P0011 set on the sibling page give timing chain 16 (14.95%), engine replacement 10 (9.35%) and VVT hardware 24 (22.43%). The two codes invert: P0011 owners name the solenoid and P0016 owners name the chain, and the engine-replacement share is 2.4 times higher here. Both sets were built with the same query, the same date range and the same exclusions, so the comparison is between like and like. SAMPLE READING, timing chain. 30 of the 49 matches were drawn at random (random.seed(3)) and read in context. THIRTY OF THIRTY are genuinely about a timing chain, guide, tensioner or a chain that jumped or skipped teeth. Observed false-positive rate: 0 of 30. No tightening was needed. SAMPLE READING, engine replacement. The pattern deliberately catches replacement that was PERFORMED, QUOTED or STATED AS REQUIRED, because a complaint often stops at the quote. 30 of the 37 matches were drawn at random (random.seed(42)) and read in full. ALL THIRTY describe a real engine replacement, rebuild or quote attached to this vehicle. One of the thirty (a 2008 Mazda CX-9) attributes the replacement to a head gasket and coolant in the oil rather than to the timing fault, which is left in the count and said out loud on the page: this number measures how often the complaint ENDS IN AN ENGINE, not how often P0016 CAUSED one. Observed false-positive rate for the intended meaning: 0 of 30; observed rate of "engine replaced for a different named cause": 1 of 30. A narrower pattern that requires the replacement to have HAPPENED ("engine was replaced", "replaced the engine") returns 5 of 162. Both numbers are published because they answer different questions: 37 is how often the conversation reached an engine, 5 is how often the owner recorded it as done at the time of writing. Published as p0016-what-owners-name.csv. ------------------------------------------------------------------------------ 6. WHO REPORTS IT, AND WHEN — INCLUDING THE 2025-2026 SURGE ------------------------------------------------------------------------------ By make and model: published in full as p0016-complaints-by-make-and-model.csv, with a manufacturer_group column and a per-model median mileage where at least one reading exists. Raw counts across vastly different sales volumes. NOT a reliability ranking. Note: the flat file spells Mercedes both "MERCEDES-BENZ" and "MERCEDES BENZ"; the two spellings are merged here and the merge is visible in the CSV. GM brands combined 89 of 162 (54.9%) GMC 39, Chevrolet 38, Cadillac 6, Buick 4, Saturn 2 Ford and Lincoln 30 (18.5%) Ford F-150 alone 21 Mercedes-Benz 19 (11.7%) E350 8, ML350 4, C230 3 Volkswagen Group 6 Mazda 5 (CX-9 5) Stellantis brands 4 Subaru 3 Kia 2 Suzuki 1 Three clusters carry the file, and each has a different shape: A. GM FULL-SIZE TRUCKS AND SUVS, 58 complaints (35.8% of the whole set). Sierra 1500 26, Silverado 1500 18, Yukon/Yukon XL 6, Tahoe 3, Escalade 4, Suburban 1. Model years 2018-2026, concentrated in 2021 (14), 2024 (12), 2020 (9) and 2019 (7). FIFTY-SEVEN OF THESE 58 REPORTS ARRIVED IN 2025 OR 2026. Twenty-six of the 58 name the 6.2L or the engine code L87; eight name recall 25V-274 or GM's own campaign numbers in the N2524940xx family. B. FORD F-150 3.5L ECOBOOST, 21 complaints, model years 2011 (7), 2012 (5), 2013 (5), 2014 (4) -- exactly the range of Ford TSB 17-0026 -- with reports arriving 2015-2023 and nothing after. Two of the 21 quote the earlier bulletin number 14-0194 by hand. C. MERCEDES-BENZ V6 AND V8, 19 complaints, 17 of them model year 2006 (9) or 2007 (8), reports arriving 2016-2025. P0017 is named alongside P0016 in 17 of the 19, and 13 of the 19 name a balance shaft, an idler gear or a sprocket. Model year and year of receipt: published as p0016-model-year-and-report-year.csv. Reports by calendar year of receipt: 4, 9, 8, 9, 22, 11, 8, 6, 8, 6, 30, 41 for 2015 through 2026, the last a PARTIAL year to 8 September. 71 of 162 (43.8%) arrived in 2025-2026. DO NOT READ THAT LINE AS A FAILURE TREND. It is dominated by one cluster: strip the GM full-size trucks out and the 2025-2026 total falls from 71 to 14. A recall that reaches 597,630 owners teaches those owners a five-character string, and some of them then type it into a federal form. The line measures attention. ------------------------------------------------------------------------------ 7. WHAT THE RECALL FILE DOES NOT CONTAIN (A PUBLISHED NEGATIVE) ------------------------------------------------------------------------------ Query: every recall row whose defect, remedy or notes text contains "P0016", case-insensitive, across the whole recall flat file. RESULT: ZERO campaigns. The same query for "timing chain" anywhere in a recall defect description returns TWO campaigns in the whole 2015-2026 file, neither on a US passenger car (a Ducati motorcycle and a Mitsubishi Fuso truck). The one US passenger-car timing chain recall this analysis found, GM 07V-519, predates the file and was retrieved from the live campaign API. The Check Engine Code Index independently reports zero warranty extensions or emissions campaigns among the 49 P0016 bulletins. This is published as an honest negative: a stretched timing chain is, with one twenty-year-old exception, a warranty-and-bulletin matter rather than a recall matter, and the page says so instead of implying a free repair might exist. ------------------------------------------------------------------------------ 8. PRIMARY DOCUMENTS OPENED FOR THIS PAGE (19 September 2026) ------------------------------------------------------------------------------ Every document below was downloaded from static.nhtsa.gov or read from NHTSA's live campaign API on 19 September 2026; every date, population, labor time and quoted phrase comes from the document itself. NHTSA recall 25V-274, Part 573 report submitted 24 April 2025, GM number N252494000. 597,630 potentially involved vehicles; 2021-2024 Cadillac Escalade and Escalade ESV, Chevrolet Silverado 1500, Suburban and Tahoe, GMC Sierra 1500, Yukon and Yukon XL with the 6.2L V8 (RPO L87), built 1 March 2021 to 31 May 2024; estimated 3 percent with the defect. Cause: "(1) rod-bearing damage from sediment on connecting rods and crankshaft-oil galleries; and (2) out of specification crankshaft dimensions and surface finish." Chronology: "GM's investigation identified 28,102 field complaints or incidents in the US potentially related to failure of the L87 engine ... of which 14,332 involved allegations of loss of propulsion," received 29 April 2021 to 3 February 2025, with 12 alleged crashes, 12 alleged injuries and 42 alleged fires. THE DEALER BULLETIN IS WHY THIS DOCUMENT MATTERS FOR P0016. Its service procedure opens: "1. Inspect/Check for DTC P0016. 1.1. If DTC P0016 is NOT SET, continue to step 2. 1.2. If DTC P0016 is set, DO NOT proceed with the rest of the service procedure, quarantine vehicle and submit labor op 9107915." Vehicles that pass get "NEW 0W-40 Oil" (dexos R, part 19432866) and a new oil fill cap; labor 0.5 hour to pass, 0.2 hour to fail. NOTE ON TWO INSTRUMENTS: the Part 573 report as submitted says 597,630 potentially involved; NHTSA's live campaign API returns 597,571 for the same campaign. Other Pulscar pages quote the API figure, so the pages stay consistent on 597,571 and record the 59-vehicle difference here rather than silently picking one. https://static.nhtsa.gov/odi/rcl/2025/RCLRPT-25V274-1598.PDF https://static.nhtsa.gov/odi/rcl/2025/RCRIT-25V274-6343.pdf Ford TSB 17-0026, 12 May 2017, NHTSA document 10099001, read in the manufacturer communications flat file (the PDF was not retrievable at the static paths tried, and that is stated rather than worked around). 2011-2014 F-150 with the 3.5L GTDI engine "built on or before 10-Oct-2014" may set P0016: "This concern may be due to a worn primary timing chain. A kit has been developed to aid in the repair process." GM 17-NA-272, August 2017, NHTSA document 10140397. 2016-2017 Cadillac CT6 and 2017 XT5 with the LGX or LGW V6, listed BY INDIVIDUAL VIN -- 28 VINs in the whole bulletin. Condition: "loud noise, malfunction indicator lamp illuminated and/or vehicle will not start" with P0016, P0017, P0018 or P0019. Cause: "a broken engine crank sprocket caused by incomplete machining. Failure of the sprocket will cause the timing chain to stop moving or break." Correction: "Replace the engine assembly," labor operation 4067490. https://static.nhtsa.gov/odi/tsbs/2017/MC-10140397-9999.pdf GM 12-06-01-009F, March 2015, NHTSA document 10113723. Timing chain service kits for the 3.6L High Feature V6 across 20 nameplates, model years 2007-2012, built after 15 July 2006, for "repairing stretched timing chains setting DTCs P0008, P0009, P0016, P0017, P0018 and/or P0019". The sentence that matters to an owner reading a quote: "Idler gears, camshaft actuators, crankshaft driven gears and timing chain guides are being replaced unnecessarily and are not worn and should not be replaced." Guides are to be replaced "only in cases where the guide wear block surface is worn down into the base friction material." https://static.nhtsa.gov/odi/tsbs/2015/MC-10113723-9999.pdf GM 23-NA-091, May 2023, NHTSA document 10236106 (replaces 19-NA-150). 2016-2023 Cadillac ATS/CT6/CTS/XT5/XT6, Chevrolet Camaro and Colorado, GMC Acadia and Canyon with the 3.6L LGX or 3.6L LGZ. "Some dealers may find DTC P0016 and/or P0018 set after an engine repair ... This will be a hard code and will usually cause a rough idle concern." Cause: the connectors for the bank 1 intake camshaft actuator solenoid and the intake actuator park lock solenoid "have been swapped. These two connectors are identical and can only be identified by wire color." https://static.nhtsa.gov/odi/tsbs/2023/MC-10236106-0001.pdf GM PIP5554A, published 7 February 2019, NHTSA document 10154276. 2014-2019 Gen 5 V6 and V8 trucks, SUVs, Camaro, Corvette and Escalade. Condition: "P0016 setting after internal engine repair." Cause: "The engine Base timing is not being set correctly," because a straight-on view of the timing marks is difficult; the bulletin teaches a method of counting chain links instead. https://static.nhtsa.gov/odi/tsbs/2019/MC-10154276-9999.pdf GM service update N252510320, dealer message dated 8 September 2025, NHTSA document 11022904 (its sibling N252510321 covers the same ground). 2023-2024 Escalade, Escalade ESV, Silverado 1500, Suburban, Tahoe, Sierra 1500, Yukon and Yukon XL: "Certain vehicles listed above will receive a software update to add to the existing Check Engine Light a Driver Information Center (DIC) warning message if DTC P0016 sets." No parts; 0.4 hour to reprogram, or over the air. The fault behind the code is untouched -- the driver is simply told about it more loudly. https://static.nhtsa.gov/odi/tsbs/2025/MC-11022904-0001.pdf Land Rover LTB00916NAS5, 18 May 2018, NHTSA document 10142778. 2013-2015 LR2, 2015-2017 Discovery Sport and 2012-2017 Range Rover Evoque, all built at Halewood, listed by VIN range. "A rattle noise may be evident from the engine compartment and/or the Malfunction Indicator Lamp (MIL) is illuminated and Diagnostic Trouble Code (DTC) P0016, P0017, P0026, or P0341 may be stored." Cause: "the 'stopper' in the intake camshaft Variable Cam Timing (VCT) actuator becoming detached." https://static.nhtsa.gov/odi/tsbs/2018/MC-10142778-9999.pdf Kia TSB ENG 170, June 2017, NHTSA document 10116121. 2017 Forte (YDm) 2.0L MPI built 15 May 2016 to 26 April 2017, MIL with P0016 "due to a stuck intake CVVT assembly". The correction is to reset the engine ECU adaptive values with the KDS scan tool -- no parts at all. https://static.nhtsa.gov/odi/tsbs/2017/MC-10116121-9999.pdf NHTSA engineering analysis EA06-009, closing resume dated 9 November 2007, read in the ODI investigation flat file. Timing chain failure causing stall while driving on 2001 Saturn L-Series 2.2L (L61) cars built November 2000 to February 2001. "GM ESTIMATED THAT 55% OF ALL TIMING CHAIN WARRANTY REPAIRS INVOLVED SWD [stall while driving] INCIDENTS. ODI'S ANALYSIS ESTIMATED APPROXIMATELY 76% OF WARRANTY REPAIRS MAY HAVE INVOLVED A SWD ... NONE OF THE VEHICLES THAT EXPERIENCING STALLS DUE TO TIMING CHAIN FAILURE COULD BE RESTARTED. IN ADDITION, APPROXIMATELY ONE-THIRD OF THE STALLS OCCURRED AT SPEEDS GREATER THAN 40 MPH AND ABOUT 20% INDICATED THAT THEY WERE NOT ABLE TO GET THE VEHICLE OUT OF TRAFFIC AFTER THE ENGINE DIED." Stall-while-driving rates were estimated at 1.7 percent (GM) and 2.3 percent (ODI) at three years in service, and 5.2 percent and 7.0 percent at six years. Closed on GM's recall. Related: defect petitions DP05-008 and DP08-002, preliminary evaluations PE06-006 and PE08-061. NHTSA recall 07V-519, report received 7 November 2007, read live on the campaign API on 19 September 2026: 20,514 2001 Saturn LS cars, "LINKS IN THE ENGINE'S TIMING CHAIN CAN SEPARATE ... THE ENGINE WILL STALL AND WILL NOT RESTART INCREASING THE RISK OF A CRASH." Dealers replaced the timing chain free of charge; the recall began 15 January 2008. https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=07V519000 NHTSA recall query RQ22-002, opened 22 July 2022, CLOSED WITHOUT ACTION 16 June 2025, closing resume read in the ODI investigation flat file. It examined whether recall 16V-907 (42,231 2016 Dodge Journey, Jeep Compass and Jeep Patriot, crankshaft or camshaft sensor connector, verified live on the campaign API) should be widened after 127 owner questionnaires from outside its scope. ODI separated two failure modes -- a faulty connector and a faulty sensor -- and found that outside the recalled population the complaints "commonly do not describe a LOMP [loss of motive power]" but "typically describe experiencing symptoms such as a MIL, Diagnostic Trouble Code (DTC) associated with crank/camshaft position sensor, and subsequent visits to a dealership." https://api.nhtsa.gov/recalls/campaignNumber?campaignNumber=16V907000 ------------------------------------------------------------------------------ 9. FILES IN THIS FOLDER ------------------------------------------------------------------------------ p0016-complaints-by-make-and-model.csv make, model, group, count, share, median miles, mileage readings p0016-what-owners-name.csv parts and contexts named, with counts p0016-model-year-and-report-year.csv model year and year of report p0016-outcomes-and-mileage.csv safety flags vs baseline, odometer bands method.txt this file ------------------------------------------------------------------------------ 10. LIMITATIONS IN ONE PLACE ------------------------------------------------------------------------------ * 162 complaints is a small set. Single-model counts in single digits are noise. * Complaints are self-reported and unaudited. They are not failure rates. * There is no denominator: no sales, registration or fleet figure is in the data. * The odometer is missing for 90 of the 162 reports. * One cluster (GM full-size trucks, 58 reports) dominates both the recent years and the engine-replacement share. Every time-series figure on the page is published with and without it. * Categorisation is one reader's judgement. The raw category definitions are printed above so any specific call can be disputed. * Keyword sets under-count. Every count here is a floor. * Selection bias runs one way: an owner whose chain was replaced under warranty at 40,000 miles rarely files a federal complaint. Nothing here can be read as "how often P0016 is a timing chain".