Measurement and bounded call
WVU CAFEE's 15 May final report, published by ICCT on 30 May, anonymized three vehicles: A used a lean-NOx trap (LNT); B and C used urea-based selective catalytic reduction (SCR). No fault codes were observed before selection. Vehicle C remains part of the comparison. VW1 study.
Call at the time: investigate possible test-dependent controls, not pronounce misconduct. Two weighted FTP-75 chassis-dynamometer means were below the study's benchmark, while road-route PEMS means for A and B were much higher. These are different tests; road means are not certification-cycle compliance verdicts.
Comparison basis
Tier 2 Bin 5 full useful life NOx: 0.07 g/mi. One mile = 1.609344 km, giving 0.043495983 g/km before rounding. The study prints 0.043 g/km. Every multiple below divides the route mean by the study's printed 0.043 g/km. VW2 Tier 2 rule; VW1 study.
Weighted FTP-75 laboratory results
CARB El Monte chassis dynamometer, g/km NOx; sigma is variation across repeated tests. Vehicle C has no laboratory result in the supplied report excerpts. VW1 pp. iii, 63-64.
| Vehicle | Mean | Sigma | Repeats | Source |
|---|---|---|---|---|
| A | 0.022 | 0.006 | 2 | VW1 study |
| B | 0.016 | 0.002 | 3 | VW1 study |
| C | missing | missing | missing | VW1 study |
Complete warm/hot-start PEMS road record
Mean and run-to-run sigma in g/km. Multiples are rounded to three decimals; missing is not zero. Two consecutive runs per route where reported. All rows: VW1 p. 65, Table 4.3.
| Vehicle | Route | Mean | Sigma | Times | Source |
|---|---|---|---|---|---|
| A | 1 highway | 0.614 | 0.283 | 14.279 | VW1 study |
| B | 1 highway | 0.344 | 0.096 | 8.000 | VW1 study |
| C | 1 highway | 0.048 | missing | 1.116 | VW1 study |
| A | 2 urban (LA) | 0.989 | 0.114 | 23.000 | VW1 study |
| B | 2 urban (LA) | 0.809 | 0.075 | 18.814 | VW1 study |
| C | 2 urban (LA) | 0.070 | 0.041 | 1.628 | VW1 study |
| A | 3 rural-up/downhill | 1.505 | 0.181 | 35.000 | VW1 study |
| B | 3 rural-up/downhill | 0.671 | 0.016 | 15.605 | VW1 study |
| C | 3 rural-up/downhill | 0.409 | 0.029 | 9.512 | VW1 study |
| A | 4 urban (San Diego) | 1.480 | 0.335 | 34.419 | VW1 study |
| B | 4 urban (San Diego) | 0.675 | 0.057 | 15.698 | VW1 study |
| C | 4 urban (San Diego) | missing | missing | missing | VW1 study |
| A | 5 urban (San Francisco) | missing | missing | missing | VW1 study |
| B | 5 urban (San Francisco) | 0.815 | missing | 18.953 | VW1 study |
| C | 5 urban (San Francisco) | 0.053 | 0.021 | 1.233 | VW1 study |
Original limits
The limited data set does not necessarily permit drawing more generalized conclusions for a specific vehicle category or after-treatment technology.VW1 study, p. iv,
Route 3 has uphill/downhill terrain. Vehicle A's Route 1 regeneration event coincided with evening stop-and-go traffic, confounding the reported 97% run difference. The original data do not identify a mechanism or establish intent; three individual vehicles cannot establish a fleet-wide rate. VW1 pp. 64-65.
Allegation, then corporate stipulation
18 September 2015: EPA's notice
EPA issued a notice of violation to Volkswagen AG, Audi AG and Volkswagen Group of America. The notice is an allegation; the searchable VW4 EPA announcement describes it; VW3 scanned notice is the notice itself.
The NOV alleges that four-cylinder Volkswagen and Audi diesel cars from model years 2009-2015 include software that circumvents EPA emissions standards for certain air pollutants.VW4 EPA announcement,
EPA described test-detecting software and reduced control effectiveness during normal driving. The agency attributed the discovery to investigations after the independent road analysis raised questions, not to the May paper identifying code:
EPA and CARB uncovered the defeat device software after independent analysis by researchers at West Virginia University, working with the International Council on Clean Transportation, a non-governmental organization, raised questions about emissions levels, and the agencies began further investigations into the issue.VW4 EPA announcement,
18 September 2015: names disclosed by ICCT
This later identification matches after-treatment systems to the original anonymized labels: A: Volkswagen Jetta, LNT; B: VW Passat, urea-SCR; C: BMW X5, SCR. These names belong only to this later column.
The three test vehicles were certified to US-EPA Tier2-Bin5 and California LEV-II ULEV emissions limits and were equipped with NOx after-treatment technologies: the Jetta was tested with a lean-NOx trap (LNT), the Passat with a urea-based selective catalytic reduction (SCR) system, and the BMW X5 with an SCR system.VW6 ICCT statement,
11 January 2017: defendant's agreement
In the Volkswagen AG Rule 11 plea agreement, the defendant stipulated to the truth of Exhibit 2's facts and corporate responsibility for the acts described there. This is a corporate plea agreement, not a court judgment at signing. The stipulated narrative describes both the on-road/dynamometer discrepancy and later additional agency testing. It does not make the mechanism knowable from the 2014 report alone.
Reveal what the defendant stipulated, beyond the road readings
The Defendant admits, agrees, and stipulates that the factual allegations set forth in Exhibit 2 (the Statement of Facts) are true and correct, that it is responsible under the laws of the United States for the acts of its employees described in Exhibit 2, and that the facts set forth in Exhibit 2 accurately reflect the Defendant's criminal conduct and provide a factual basis for the guilty plea.VW5 agreement, section 1.E,
Following the ICCT study, CARB, in coordination with the EPA, attempted to work with VW to determine the cause for the higher NOx emissions in the 2.0 Liter Subject Vehicles when being driven on the road as opposed to on the dynamometer undergoing standard emissions test cycles.VW5 Exhibit 2, paragraph 53,
The defendant's stipulated account says explanations and fixes did not reveal the underlying reason:
without revealing the underlying reason-the existence of software designed to detect, evade and defeat U.S. emissions tests.VW5 Exhibit 2, paragraph 55,
These are the defendant's later stipulations, not deductions of software or intent from the original PEMS measurements.
Later source ledger
- VW3 EPA notice, 18 September 2015, scanned original.
- VW4 EPA announcement, 18 September 2015.
- VW6 ICCT statement, 18 September 2015.
- VW5 corporate agreement, 11 January 2017, section 1.E and Exhibit 2, paragraphs 52-55.