UN GTR 21 (DEVP) — 원문·번역 대역

1. Passenger vehicles are commonly assigned a vehicle power rating, which is useful for comparing the performance of different vehicles. Vehicle power rating has also been used for other purposes such as vehicle classification, customer information, insurance, and taxation.

2. Historically, almost every passenger vehicle produced for the consumer market has been powered exclusively by an internal combustion engine (ICE). The vehicle power rating assigned to these conventional vehicles has customarily been the same as the rated power of the engine, as determined by an engine bench test. This is a convenient way to assign a power rating to a vehicle, because the engine power rating may then be applied to any vehicle that uses the same engine.

3. As a measure of real-world vehicle performance, this traditional measure is imperfect, since it does not account for the power lost in the drivetrain between the engine and the road. However, it has become well established and is generally accepted as a useful metric, in part because conventional vehicles have only one engine, and its full rated power is typically available for propulsion.

4. Today, electrified vehicles such as hybrid electric vehicles (HEVs) and pure electric vehicles (PEVs) with multiple drive motors represent an increasing share of the market. A vehicle power rating is not as easy to assign to these vehicles because they combine more than one propulsion source, such as an engine and an electric machine, or multiple electric machines.

5. For these vehicles, the available power depends on how the control system combines the power of each propulsion source when the driver demands maximum power. While it may seem that this would simply be the sum of the rated power of each component, this is not necessarily valid in practice. It will result in an overestimate if, for example, the electric machine is limited by the available battery power, or if the control system limits or reassigns some of the nominal capacity, such as to maintain traction or charge the battery.

6. Owing to the pressing need to reduce emissions of greenhouse gases (GHG) and other air pollutants, the market share of electrified vehicles is expected to grow in the future. This intensifies the need for a standard method for assigning a vehicle power rating to electrified vehicles.

7. Electrified vehicles and conventional vehicles are likely to coexist in the market for some time. Many existing regulations and procedures, such as WLTP, apply to both conventional and electrified vehicles, and require a power rating as an input. In order to be used equitably for such purposes, a power rating for electrified vehicles should be qualitatively and quantitatively comparable with the traditional engine-based power ratings of conventional vehicles. B. Technical background

8. Developing a test procedure and system power rating that fits the requirements presents two primary technical challenges: (a) The first challenge is to identify a reliable and repeatable way to command a vehicle to deliver maximum power in a laboratory setting. (b) The second challenge is to identify a comparable and valid basis for the system power rating and to identify the measurements and calculations necessary to produce it. (a) Commanding maximum power

9. As part of their standards development efforts, SAE and ISO studied ways to elicit maximum power in a laboratory setting (more details about the Informal Working Group (IWG) on Electric Vehicle and the Environment (EVE), SAE and ISO activities related to the topic of power measurement of electrified powertrain can be found in the technical report of this UN GTR). This resulted in identification of a reliable and repeatable method to do this by use of the fixed-speed setting of a chassis dynamometer. The condition of maximum power is determined by driving the vehicle on the dynamometer at a series of fixed dynamometer speeds to find the maximum brake power of the dynamometer that the vehicle is able to run against. At each speed, the accelerator is rapidly and fully depressed for at least 10 seconds. The speed at which the dynamometer records the highest power is recorded. The system power is then determined at this fixed dynamometer speed. (b) Basis and measurements

10. In early discussions, the IWG on EVE discussed a number of conceptually simple measurement bases for electrified vehicle power.

11. One very simple basis would simply measure the peak power delivered to the wheels. This would be compatible with all electrified vehicles regardless of their powertrain architecture. If also extended to conventional vehicles, it would rate all vehicles on a directly comparable basis, and would represent real-world power more effectively than the traditional measure because it includes the effect of losses in the drivetrain. However, for the same reason, this would result in power ratings that are not comparable to the traditional measure, which continues to be used in many applications.

12. Another simple approach would measure the peak power delivered to the wheels and then adjust it by an assumed transmission efficiency. This approach recognizes that an engine-based power rating, in theory, should be identical to the peak power delivered to the wheels divided by the mechanical conversion efficiency of the drivetrain (e.g. gearbox or transmission). By extension, a highly comparable power rating for an electrified vehicle could be determined by measuring the peak power delivered to the wheels and dividing by a typical (conventional) drivetrain efficiency at peak load, perhaps 90 to 95 percent. However, it was not clear that this approach would represent all hybrid powertrains equally, nor that a single assumed drivetrain efficiency would represent all comparison vehicles equally.

13. Another possibility would sum the power of the engine with the measured power output of the battery. Many hybrid vehicles operate the engine at full throttle when the driver demands maximum power, meaning that engine power can be estimated from engine speed by reference to a full load power curve. Battery power is also reasonably simple to measure, and measuring at the battery avoids the need to instrument individual inverters or motors. However, it would neglect electrical conversion losses in the latter, and so might tend to produce optimistic results for highly electrified powertrains.

14. Recognizing that these relatively simple methods vary in their comparability and fairness, the IWG on EVE sought to identify a more sophisticated approach.

15. Conceptually, a comparable and fair rating would be based on the power that passes through the powertrain at a point that is mechanically analogous to the output shaft of a conventional engine (as opposed to the wheels or the battery, where the losses would be different). Intuitively, this point would include the mechanical output shafts of any propulsion energy converters (i.e. engine and electric machines) that contribute propulsion energy when the driver commands maximum power.

16. As an example, Figure 1 illustrates a typical P2 hybrid configuration, in which ICE power and electric motor power is mechanically combined on a single shaft. It identifies two "reference points," R1 and R2, which together are mechanically analogous to the power output of the engine in a conventional vehicle. That is, they represent where the mechanical power that drives the wheels is first produced from stored energy. The goal would be to determine the sum of the mechanical power passing through R1 and R2 when the vehicle is producing maximum power. Figure 1 Example of reference points for system power determination

17. In theory, the most direct approach to measure the power at R1 and R2 would be to instrument the corresponding shafts with torque and speed meters. However, this requires invasive instrumentation, may not be possible in some cases, and is unlikely to be practical in a type approval context.

18. A more practical approach would measure power flow at other points in the powertrain that are easier to instrument, and estimate the power at reference points R1 and R2 by accounting for the losses between the measuring points and the reference points. As shown in Figure 2, the measuring points could either be upstream or downstream of the reference points. An option for measuring upstream (option 1) might include measuring engine speed and converting it to the mechanical power output at R1, and measuring REESS power output and converting it to the power at R2 by accounting for electrical conversion losses. Options for measuring downstream (option 2) might include measuring the power delivered to the axle by means of wheel torque and speed sensors or a hub dynamometer, and then determining the sum of R1 and R2 by accounting for mechanical conversion losses in the drivetrain. Figure 2 Possible measurement points to estimate power at R1 and R2 for parallel P2 hybrid Note: measurement point for option 2 represents both axle shafts.

19. Electrified powertrains vary widely, and can include power flow paths that are much more complex than those depicted here. However, once the reference points are identified, it should be possible to estimate the power at the reference points by taking appropriate measurements when the vehicle is generating maximum power, and accounting for the losses between the measurement points and the reference points using component test data or sound engineering judgement.

20. It should be noted that the traditional engine-based metric does not perfectly represent the road power available to the driver, because it neglects losses in the transmission. This also makes it imprecise, in that the road power may vary significantly from one vehicle model to another due to differences in drivetrain losses.

21. Engine power ratings are also somewhat imprecise. For example, UN Regulation No. 85 allows the declared power value for a production engine to vary by ± 2 percent from the certification test result, and by ± 5 percent for conformity of production.

22. A system power metric for electrified vehicles might therefore be held to a similar level of accuracy and precision.

23. The IWG on EVE received presentations from experts with several organizations that were studying the problem of hybrid system power determination. (a) SAE J2908

24. The SAE J2908 Task Force led by Argonne National Laboratory (ANL) started its project in November 2014. Three primary methods of determining HEV system power were initially investigated (referred to here as Method 1, Method 2, and Method 3).

25. SAE Method 1 was the sum of engine power (estimated from bench test results) and measured DC power from the battery (neglecting electrical conversion losses in the inverter and electric machines). SAE Method 2 was the sum of estimated shaft powers from the engine and the electric machines (determined from bench test results and onboard data, respectively). SAE Method 3 was the measured power at the axle or wheel.

26. The IWG on EVE agreed with the characterization of these three primary methods as reasonable approaches to measure system power. However, the three methods varied in terms of how well the measure could be compared to the traditional power ratings for conventional vehicles, and in terms of the ability to verify a reported value. Method 1 was conceptually similar to the conventional engine-based rating and would be straightforward to verify by measurement, but neglected some losses. Method 2 was most comparable to the conventional rating, but would impose the highest burden of instrumentation to verify. Method 3 would be easily verifiable by dynamometer testing, but because a wheel power measurement accounts for losses in the drivetrain, it would not be as comparable to the conventional rating, which does not. (b) KATRI standard

27. KATRI started a research project in July 2013 with the aim of developing a national standard for the determination of a representative power for (Non) Off-Vehicle Charge ((N)OVC)-HEVs and PEVs with in-wheel motors. It was completed in June 2015. Nominal rating and system power tests were studied using a powertrain dynamometer or a chassis dynamometer with added instrumentation. The definition of hybrid system power follows the same approach as SAE Method 1, namely that it involves a simple addition of the rated engine power and the electric power of the battery. The engine power is the rated power according to UN Regulation No.

28. ISO conducted a project under New Work Item Proposal (NWIP) N3477 proposed by the Japan Automobile Research Institute (JARI), approved in June 2015. It started as a formal project of ISO/ TC22/SC37/WG02 and was finalized as ISO Standard 20762 in 2018.

29. The ISO method includes two alternative test procedures, referred to as test procedure 1 (TP1) and test procedure 2 (TP2). Note: measurement point for TP2 represents both axle shafts.

31. TP1 is similar to SAE Method 1, but additionally accounts for electrical conversion losses. Total power is the sum of estimated engine power and estimated motor power. Engine power is the rated power by ISO 1585 (or UN Regulation No. 85) at the observed operating point. Motor power is based on measured REESS power, adjusted by a factor (known as K, with a default value of 0.85) that represents combined efficiency of the inverter(s) and electric machine(s). (Electrical power to the accessories is also estimated or measured and deducted from the REESS power.) Figure 4 illustrates how total power is modelled under TP1. Figure 4 TP1 as sum of estimated engine power and estimated motor power

32. TP2 is similar to SAE Method

33. It could be said that TP1 and TP2 provide the flexibility in measurement options provided by SAE Method 1 and 3, while the inclusion of the adjustment factors K and ηgb result in a metric more like that of SAE Method 2, which is most comparable to the traditional measure.

34. In both TP1 and TP2, power is measured when the hybrid system as a whole delivers maximum power on a dynamometer running at a fixed speed. If not provided by the manufacturer, the fixed speed at which maximum power is delivered is determined by carrying out a series of test runs while driving the vehicle on the dynamometer at a series of fixed speeds to find the maximum brake power of the dynamometer that the vehicle is able to run against. At each speed, the accelerator is rapidly and fully depressed for at least 10 seconds.

35. As shown in Figure 6, the tests result in a power-versus-speed curve that helps to identify the fixed dynamometer speed at which maximum power is generated. If necessary, the evaluation is continued with smaller speed steps near the peak of the curve until the speed of the peak power is accurately identified. The power test is then performed at this fixed speed. Figure 6 Identification of speed of maximum vehicle power

36. Calculations are then performed to determine the system power according to TP1 or TP2. As shown in Figure 7, a "peak" power is defined as the maximum value of a 2-second moving average of the total power over a 10 second window beginning at the start of maximum accelerator command, and a "sustained" power is the average total power between the 8th and 10th seconds. Figure 7 Definition of peak and sustained power

37. The IWG on EVE recognized that the ISO method showed good comparability, flexibility, and verifiability. At the 22nd meeting of the IWG on EVE, the contracting parties reached consensus that the ISO approach presented the best option as a basis to fulfill the needs of the mandate.

38. The IWG on EVE then turned attention to aligning and integrating the ISO method with UN GTR No. 15, or developing a new GTR. There was some debate as to whether the GTR should select only one of the ISO test procedures (TP1 or TP2) or retain both options. It was generally decided that retaining both would be preferable because it would accommodate variations in vehicle instrumentation possibilities and differing laboratory capabilities or preferences.

39. The IWG on EVE recognized that retention of both procedures meant that differences between the two test results should be minimized in order to prevent inconsistent results and the opportunity for selective reporting (or "cherry picking").

40. In designing and validating the ISO method, the ISO committee placed strong emphasis on its practicability. Testing at the Japan Automotive Research Institute (JARI) indicated that the procedures delivered equivalent results for a variety of HEVs, although TP2 was thought to show somewhat greater variability than TP1. Discussion in the IWG suggested that the relative variability may be the result of TP2 being based entirely on measured data, while a large component of TP1 relies on a fixed value for engine power obtained from the UN Regulation No. 85 rated power. If so, then the relative variability may be a natural outcome of differences in the procedures.

41. The IWG on EVE recognized that additional validation testing would be necessary to assess this and other potential sources of variability, and also to validate the ability of the aligned ISO method to fulfil the specific needs of a regulatory application.

42. Several contracting parties volunteered to perform validation testing, including Environment and Climate Change Canada (ECCC), Joint Research Centre (JRC), U. S. Environmental Protection Agency (EPA), and KATRI.

43. A first phase of the validation program was initiated at the April 2018 IWG on EVE meeting in Tokyo. Japan reviewed the testing performed on three HEVs in conjunction with development of the ISO standard in 2016. A matrix of additional HEVs that were available for testing was compiled. US EPA offered to perform testing of a Belted Alternator Starter (BAS) hybrid and a power split PHEV. Canada offered to perform testing of a later generation power split HEV, a P2 hybrid, and a two-motor PEV. KATRI offered to perform testing on a P2 hybrid. JRC offered to perform testing on two parallel hybrid vehicles provided by representatives from Volvo and Hyundai.

44. Japan arranged for consultation with the engineer who performed the ISO validation tests in Japan. A detailed technical report on this testing had been prepared in Japanese. Canada agreed to arrange for translation of the report into English. JRC scheduled an initial round of testing at the facilities in Ispra, Italy in 2018, which was attended by representatives from USA and Japan as well as technical support personnel from Volvo and Hyundai.

45. Due to the short time frame available, and the knowledge that the ISO committee had already performed significant validation, the validation testing focused primarily on practicability of the procedure as currently written, and the effect of default assumptions and available flexibilities on the consistency of the results. To save time, testing was limited to vehicles that were readily available at the participating test labs and calculations were performed using the specified default values for K (later renamed K1) and ηgb (renamed K2). In some cases, measurements were gathered from onboard systems rather than instrumentation due to resource constraints. While the measurements were believed to be sufficiently accurate, it was not always possible to validate onboard measurements for accuracy.

46. The results of the first phase of validation revealed significant and unexpected differences between the results of TP1 and TP2 for many of the vehicles tested. Accordingly, the work of the IWG began to focus on identifying the sources of these differences, their implications, and how to reduce or eliminate them.

47. The IWG on EVE identified several potential causes for the observed differences: (a) Variation in accuracy of default values for K1 and K2 as applied to specific vehicle models. (b) Uncertainty in accuracy of measurements and measurement options. (c) Variation in power of production engines compared to UN Regulation No. 85 test results. (d) Influence of powertrain architecture on necessary measurements to perform TP1 or TP2 in an equivalent manner. (a) Default values for K1 and K2

48. For a given powertrain architecture and vehicle model, the relative accuracy of the fixed default values for K1 and K2 are likely to vary, leading to differences in the accuracy with which each TP accounts for losses, and thereby leading to a difference in the results.

49. In particular, the default K1 value of 0.85 sometimes appeared to produce lower power ratings for TP1, depending on the fraction of total power contributed by electricity. For one vehicle that was propelled entirely by electrical power, the power rating delivered by TP1 was smaller than the power measured at the wheels (which would erroneously suggest a drivetrain efficiency greater than 100 percent). Modifying the K1 value to a different value that was still consistent with the powertrain design made the result much closer to that of TP2.

50. For some powertrain architectures, the applicable default K2 factor for TP2 was unclear. Two of the test laboratories independently chose to employ different K2 values for an architecture that included series and parallel elements.

51. It was anticipated that the predefined list of default K2 factors may be insufficient to represent potential architectures that may emerge in the future. In particular, Japan pointed out that it is uncertain whether the default value for K2 would apply to different variations in power split hybrid architectures. (b) Accuracy of measurements

52. Some of the validation tests relied on TP1 measurements that were based on onboard network data that could not be verified because physical instrumentation for current and voltage was not available. While believed to be accurate, any inaccuracy could have contributed to the difference between TP1 and TP2.

53. Measurements for TP2 were taken from dynamometer rollers and therefore included tire losses. While the test procedure permitted the use of roller data if tire losses were accounted for, it did not specify a method for determining tire losses. Evidence of tire slippage was observed, which may have introduced additional unaccounted losses. (c) Variability of UN Regulation No. 85 engine power

54. TP1 may be affected by allowable variation in engine power from UN Regulation No. 85 test results. According to Section 5.4 of UN Regulation No. 85 (Interpretation of results), the declared power output of production engines certified under UN Regulation No. 85 is permitted to vary by ± 2 percent from the test result, suggesting that some error is possible even if the measured engine speed and intake manifold pressure match perfectly with those reported in UN Regulation No.

55. Further, TP1’s estimation of engine power based on measured speed relies on the assumption that the engine is operating at its maximum power for that speed, and that the power can be accurately reconstructed by reference to engine test results (e.g. UN Regulation No. 85). Measurements of intake manifold pressure and fuel flow rate are compared to the engine test result to verify that the engine operating state is consistent with maximum power. However, the test procedure did not specify the permissible variation, leading to uncertainty in the engine power portion of TP1.

56. Some experts noted that intake manifold pressure is not highly sensitive to power output at the constant engine speed that results from the procedure, and therefore it is not highly effective at confirming the result. It was recommended that measurement of fuel flow rate also be required for verification of UN Regulation No. 85 engine power. (d) Influence of powertrain architecture

57. ISO 20762 does not mention the concept of reference points, although reference points are implied by the details of the procedure. When the concept of reference points was introduced and applied rigorously, it was found that for some powertrain architectures, the then-prescribed calculations for TP1 and TP2 may have been estimating power at slightly different reference points, leading to variation between the results.

58. As shown in Figure 8, both TP1 and TP2 apply well to a parallel P2 HEV. Here, the system power is the sum of the power at R1 and R2. The K1 and K2 factors represent the conversion efficiencies of simple component combinations, and so are relatively simple to determine and verify. TP1 determines engine power at R1 by reference to speed and UN Regulation No. 85 results, and determines the power at R2 by measuring power from the REESS (subtracting accessory power) and applying the K1 efficiency factor. Alternatively, TP2 determines the sum of the power at R1 and R2 by measuring power at the axle shafts and applying K2. If the applicable measurements and K factors are equally accurate, then for this powertrain architecture, TP1 and TP2 should always deliver the same answer for the sum of R1 and R2. Note: measurement point for TP2 represents both axle shafts.

59. However, in the case of some other architectures, the specified measurements for TP1 or TP2 may be difficult to convert to a common reference point.

60. As shown in Figure 9, the Toyota Hybrid System (THS) utilizes a planetary gear set with multiple inputs and outputs. Under maximum power demand, engine power enters through the planet gear carrier (P), then is split between the ring gear (where it goes directly to the wheels) and the sun gear S (where it enters a series path that eventually delivers additional torque to the ring gear for delivery to the wheels). Figure 9 Power split hybrid, ambiguous under TP2 P = planet carrier and gears; S = sun gear; Ring = ring gear Note: measurement point for TP2 represents both axle shafts.

61. With careful consideration, reference points that are most comparable to a conventional vehicle can be identified. Reference point R1 represents where mechanical power from the engine is first produced. From here, the engine power splits to the series path and the direct-to-wheels path, which together may be considered as a sort of electromechanical transmission, and therefore, as with the transmission of a conventional vehicle, is not subject to further accounting.

62. Another reference point must be established to account for the contribution of the REESS. REESS power is first produced as mechanical power at the output shaft of motorgenerator MG; however, at this point it has been combined with power contributed by the engine series path (which is already accounted for via R1). To prevent double counting, the second reference point is therefore called R2REESS, and represents the portion of MG power that is attributable to the REESS.

63. TP1 is straightforward for this architecture. The power at R1 is determined from UN Regulation No. 85 results, and R2REESS is the measured REESS power multiplied by K1 (where K1 is the electrical conversion efficiency of the total power flow through Inv1 and MG). System power is the sum of R1 and R2REESS.

64. TP2 is not as straightforward here. TP2 relies on a measure of total power at the axle shafts or wheel hubs, to which it seeks to apply a K2 efficiency factor to account for gearbox losses. But here, the power has arrived via two different paths from the engine, and a third path from the REESS, all of which have experienced different conversion efficiency. The combined power measurement at the axle does not identify the share of power along each path, so there is not enough information to reconstruct the power at R1 and R2REESS even if the conversion efficiency of each path is known.

65. Another option might be to compute (R1+R2REESS) rather than each individually. This would require a "net" K2 factor that accounts for the total losses along all three paths. If all three paths had the same conversion efficiency, it would not be necessary to know the power along each path. But that is not the case here. While the manufacturer might be able to experimentally determine a "net" K2, it would not be possible to verify using the data that is collected by TP2. If the K2 factor were to represent anything other than this "net" factor, such as for example just the efficiency of the mechanical direct drive path, then it would not be reconstructing the power at either of the designated reference points.

66. This is another way of saying that the original versions of TP1 and TP2, when applied to a power split hybrid, each determine the power at slightly different reference points. When considered individually, either of the results might be reasonable as a system power rating. However, they cannot be expected to be the same if they refer to different reference points.

67. This situation is seen more clearly in Figure 10, for a pure series hybrid. As before, the reference points are where mechanical power is first produced, at R1 and R2REESS. TP1 would determine the mechanical power from the engine (at R1) and the REESS contribution at motor MG (at R2REESS). In contrast, TP2 would measure the power at the axle shafts and apply a K2 factor to account for losses in the gearbox and differential, thereby reaching a different reference point (here called R2TOT) and reporting that as the system power. The power at R2TOT is bound to be different than at (R1 + R2REESS). Further, RTOT is inconsistent as a reference point because it is not a point where mechanical power is first produced. Figure 10 Inconsistent reference points for TP1 and TP2 for pure series HEV Note: measurement point for TP2 represents both axle shafts.

68. Further, as a side effect, here the power measured by TP2 at R2TOT will always be lower than for TP1, because the power at R2TOT has been reduced by losses in the electrical Note: measurement point for TP2 represents both axle shafts.

73. Figure 12 shows an example HEV with two powered axles. Here a four-wheel-drive dynamometer would be needed, and the power measured at each axle separately. The reference points on the first (right) axle are marked R1 and R2, and on the second (left) axle, R3. TP2 is straightforward for each axle (although it does require a unique K2 factor for each axle). TP1 can determine R1, R2, and R3 if the electrical measurement points include the inputs to each inverter (Inv1 and Inv2) and factors K1(1) and K1(2) are provided. Alternatively, TP1 can determine R1 and the sum (R2+R3) if the electrical measurement is at the REESS and the conversion efficiency of the two electrical paths can be combined or are the same. Figure 12 Vehicle with two powered axles Note: measurement points for TP2 represent both axle shafts.

74. However, as shown in Figure 13, a small change to the configuration makes it very difficult to apply TP2. Here MG2 might represent a pair of wheel hub motor(s) which now contribute to powering the first axle. The power flow from the wheel hub motors at R3 is likely to experience a very high efficiency K2(2), while those entering the gearbox/differential from (R1+R2) experience a probably lower efficiency K2(1). Because TP2 measures only the combined power, at the axle, it is not possible to apply both K factors to the portion they represent. Figure 13 Configuration with difficulty for TP2 Note: measurement point for option 2 represents both axle shafts.

75. The applicability of TP1 and TP2 can depend not only on the physical configuration of the powertrain, but also on the selected driving mode. Figure 14 and Figure 15 show two high-power modes of the Generation 2 Chevy Volt powertrain, one for a pure electric chargedepleting (CD) mode and another for a blended charge-sustaining (CS) mode.

76. In CD mode (Figure 14), both TP1 and TP2 can be performed (with certain assumptions). TP1 can determine both R1 and R2, assuming that the power into each inverter is measured, or the sum (R1+R2) if power from the REESS is measured and the conversion efficiency of both electrical conversion paths is the same and can thus be combined. TP2 can determine the sum (R1+R2) from the power measured at the axle, assuming that the efficiency of each sun-to-planet (S, P) gear path is the same. Figure 14 Volt Gen 2 charge-depleting Mode 2 (CD2)

77. However, in CS mode (Figure 15), the power flow paths are different. TP1 can still determine R1 and R2 from engine and REESS measurements. But in order for TP2 to determine the sum (R1+R2) as before, the efficiency of the Ring-to-planet and Sun-to-planet gear paths must be similar enough to be combined. Otherwise, the relative power contributed by the engine and the motor would be required, and it is not collected. Figure 15 Volt Gen 2 charge sustaining mode 2 (CS2)

78. At the 30th IWG on EVE meeting, the IWG requested that experts from VDA (German Association of the Automotive Industry) who were involved with development of the ISO procedure provide additional input on the observed differences between the results of TP1 and TP2. VDA delivered a presentation at the 31st IWG on EVE addressing this topic and provided recommendations for the second phase of validation testing.

79. The VDA experts acknowledged that some of the deviation could be the result of fixed, default K1 and K2 factors, but felt that it was also important to verify that the measurement requirements and accuracies described in ISO 20762 are followed.

80. VDA also stated that TP1 and TP2 can be expected to give the same result for parallel hybrids, which is consistent with the discussion in the previous paragraphs.

81. For pure series or mixed (power split) hybrids, VDA stated that TP1 will always give a higher result than TP2 because TP1 does not account for electrical conversion losses in the series portion. This observation has now been explained by the difference in the reference points implied by TP1 and TP2 for power split and pure series hybrids, as discussed in the previous paragraphs. Defining the reference points as depicted in Figure 9 addresses this concern, and means that TP2 becomes not applicable to this powertrain.

82. The IWG on EVE recognized that the need to reconcile TP1 and TP2 was a significant outstanding issue for the completion of the GTR. At the 30th meeting of the IWG on EVE in Stockholm, the IWG considered several options for completing the GTR.

83. One possibility was to accept the difference between TP1 and TP2, and add interpretive text to the GTR to help users understand the difference. This would maintain the flexibility of the procedure, minimize divergence from ISO 20762, and reduce the likelihood that the difference could be misunderstood or deliberately misused. This option found little support.

84. Another possibility was to eliminate the difference by modifying the GTR to define only a single possible result, rather than two. This might be done by any of: (a) Including only TP1 or TP2 in the GTR; (b) Requiring both TP1 and TP2, and reporting the average, the lower, or the higher of the two; (c) Retaining the nominal choice of TP1 or TP2, but validating the result by performing the other TP as a consistency check; (d) Specifying TP1 for some HEV architectures and TP2 for others.

85. (a) The IWG was reluctant to eliminate either TP1 or TP2 entirely, due in part to the flexibility it affords, and preferences among members for one or the other procedure.

86. (b, c) The IWG was reluctant to require both TPs to be performed because this would increase the test burden. Also, it was noted that the best choice among an average, lower, or higher of the two results would depend on the intended purpose of the measure. For downscaling and classification under WLTP, selecting the higher figure might be preferable because it would prevent excessive downscaling. But for customer information, the lower figure might be preferable to prevent exaggerating the available power. It was unclear if there was a valid technical justification for selecting either figure, or an average of the two, when it remained uncertain which result is most accurate for a given vehicle.

87. (d) The IWG remained open to the possibility of assigning TP1 and TP2 to specific powertrain types, given a clear technical justification.

88. A final possibility was to modify the procedure to minimize the difference between TP1 and TP2 as much as possible.

89. Because the problem is essentially one of physics, it should be possible to define TP1 and TP2 so that they deliver comparable results in all cases, if the following is true: (a) the power flows in the vehicle are correctly understood, (b) the reference points are correctly identified and consistent under both TP1 and TP2, and (c) the measurements and K factors are sufficiently accurate to estimate the power at the reference points.

90. The question is to what degree the procedures for TP1 and TP2 can provide for this outcome while remaining practical to implement. For example, if successfully applying TP1 sometimes requires instrumentation of several inverter inputs rather than only the REESS output, or if successfully applying TP2 requires knowledge of relative power flows that are not measurable at the axle, the instrumentation burden may become prohibitive.

91. At the 30th and 31st IWG on EVE meetings it was generally agreed that the difference between TP1 and TP2 should be reduced as much as possible by modifying the procedures, and that limiting certain architectures to TP1 or TP2 could also be considered. Several proposed modifications were identified to be evaluated in a second phase of validation testing.

92. The IWG reached consensus on several proposed modifications to reduce the difference between TP1 and TP2: (a) The option to use default K factors was replaced with a requirement that the manufacturer provide accurate and verifiable K factors specific to the vehicle under test. (b) The option to conduct TP2 using chassis dynamometer roller data was removed, in favor of axle or wheel hub instrumentation for torque and speed, or a hub dynamometer. (c) The procedure was clarified to require that current and voltage, if obtained from onboard systems, must be shown to be accurate (TP1).

93. The drafting group also proposed several changes to be trialed in the second validation phase: (a) To reduce the possibility of variation, five repetitions of the power test are conducted and an average taken of the last four results (see paragraph 6.8.7.). (b) Applicability guidelines were added to determine the permissible application of TP1 and TP2 based on aspects of the power flows between the measurement points and the reference points, and any need for additional instrumentation to enable one or the other TP (see paragraph 6.1.3.). (c) A requirement was added for the manufacturer to document the flow of propulsion power through the powertrain of the vehicle during the maximum power condition, the proposed measurement points and reference points, and applicable K factors for TP1 or TP2 (see paragraph 6.1.1.1.). (d) The term "reference point" was introduced and defined. Guidelines for identifying reference points are provided in Annex 1.

94. The new requirement that K factors be furnished by the manufacturer means that it must be possible for the manufacturer to determine the relevant K factor, and for a third party to verify it by a standard method.

95. The IWG considered that for TP1, test standards exist for the measurement of inverter and motor efficiency (K1), which could be used by the manufacturer to derive the K1 factor as well as by a third party to verify it. However, no similar test standard exists for gearbox efficiency (K2).

96. VDA was asked to provide a recommendation for a standard method for determining K2 for TP2. VDA recommended that any of various engineering methods could be employed, based on measurement of power in and power out on a test bench, and dividing output power by input power.

97. The IWG also considered a proposal that a K2 factor might be determined (or verified) by performing TP1 using a known accurate K1 factor, and then solving for K2 by setting the result of TP1 equal to the result of TP2. A similar tactic might also be usable for internal validation of a test result. This approach was to be further evaluated with data from the second phase of validation.

98. The test laboratories were requested to implement a second phase of validation testing, with the following changes to the test program: (a) Conduct TP2 with torque and speed data from torque and speed sensors rather than dynamometer roller data. (b) Conduct TP1 with current and voltage data collected from current and voltage instrumentation, in addition to onboard data. (c) If more than one electrical power path is present downstream of the battery, then instrument the inputs to each inverter (if possible). (d) Seek measurements of electrical power to non-propulsion accessories. (e) Improve precision of wheel speed and dynamometer roller speed to identify presence of wheel slippage. (f) If significant wheel slippage is observed, add weight to the vehicle to eliminate it, particularly if slippage might affect the shifting or other behavior of the vehicle.

99. In most cases, K factors were not expected to be available. Outside of a type approval or certification context, manufacturers are unlikely to have suitable data already prepared and little incentive to produce it. Even if K factors were provided, their usefulness in validating the procedure would be limited unless they could be independently verified (which was not within the scope of the program). Instead, the results were to be evaluated by considering the ability for reasonable K factors to make the results of each TP consistent with each other.

100. For the second phase of validation, ECCC tested: a 2018 BMW 530e (OVC-HEV), a 2016 Chevrolet Volt (OVC-HEV), a 2018 Toyota Prius Prime (OVC-HEV), and a 2009 Saturn Vue (mild BAS NOVC-HEV). JRC expressed an intention of testing two additional vehicles, and as of Autumn 2019 were continuing efforts to procure suitable vehicles and provide them with necessary instrumentation. US EPA had intended to test two additional vehicles, but damage to one of the vehicles, and an unexpected difficulty with the funding mechanism for contract work necessary to instrument the vehicles, made it impossible for EPA to participate in the second phase.

101. Results of the second phase began to become available in late 2019 and continued to be produced through March 2020. At an interim IWG on EVE teleconference on 12 December 2019, ECCC provided draft reports for the 2018 BMW 530e and the 2016 Chevy Volt, followed by final reports in March 2020. A report for the Saturn Vue was delivered in February 2020. As of March 2020 a draft report for the Prius Prime is awaiting completion.

102. JRC provided test results for hub dyno testing and is progressing to provide results of wheel torque measurements on the same vehicle.

103. Throughout the test program, ECCC encountered difficulty obtaining UN Regulation No. 85 engine test results applicable to the vehicles tested. UN Regulation No. 85 results were obtained for the Toyota Prius Prime in January 2020, and for the European version of the BMW 530e in February 2020 (however, the vehicle tested was a North America vehicle for which the engine has a different torque specification). Because the Chevy Volt and the Saturn Vue are not EU-spec vehicles, UN Regulation No85 data was not available for these vehicles. For these reasons, TP1 could not be performed for these in exactly the manner prescribed.

104. As for TP2 results, ECCC found that the torque and speed measurement devices gave inconsistent results and in some cases malfunctioned. There is significant doubt as to whether the TP2 results are valid due to these difficulties.

105. Although a direct comparison between TP1 and TP2 was therefore not possible in many cases, the second phase of validation revealed valuable recommendations regarding the practicability of the procedure and recommendations for improvement.

106. Additionally, late results from JRC testing with a hub dynamometer have confirmed good agreement between TP1 and TP2 for a P2 hybrid configuration. Analysis of the data will continue to further validate this conclusion and for consideration in the development of future versions of this GTR. C. Technical rationale and justification Section C.1 describes the technical justification for the major specific differences between the procedure described in this GTR and the ISO 20762 procedure on which it was based. Section C.2 provides additional discussion of the basis upon which the IWG on EVE recommends the procedure as a whole.

107. A primary anticipated use for the test procedure is for determining a system power for the purpose of classification and downscaling under the WLTP test procedure defined in UN GTR No.

108. ISO 20762 allows for K factors to be provided by the manufacturer. It also provides default K factors that could be used as needed. The IWG on EVE noted that no fixed default K factor could be expected to be equally accurate for all vehicles, and so the use of default factors could contribute to variation between TP1 and TP2.

109. Unlike ISO 20762, this GTR is likely to be applied in the context of type approval or certification. In this context, it is likely that there will be sufficient manufacturer cooperation to prevent the need to assume a default K factor.

110. This GTR therefore requires the manufacturer to provide verifiable K factor(s) in all cases, as described at paragraph 6.1.1.2. Determination and verification of the provided K factor(s) can be performed through applicable test standards or other methods as described in paragraph 6.1.1.2. (c) TP2 to utilize torque and speed sensors or hub dynamometer

111. ISO 20762 specified that measurement of torque and speed for TP2 may be acquired by use of torque and speed sensors attached to the axle shafts or wheel hubs, or by dynamometer measurements of speed and torque delivered to the dynamometer rollers. In the latter case, losses in the tires are to be accounted for. A specific method for determining the losses is not provided.

112. The IWG found that accounting for tire losses may introduce uncertainties specific to TP2. Accounting for rolling resistance requires that the rolling resistance coefficient (RRC) and the normal force on the tires both be known. RRC is not always known with high accuracy. When installed on a dynamometer, the normal force may be uncertain due to the effect of the tie down method (usually tensioned straps or chains, or rigid restraints). Tire slippage under maximum power may be difficult to eliminate, and can add losses that are difficult to quantify.

113. The GTR therefore removes the option for dynamometer roller measurements for TP2, and adds a new option to use a hub dynamometer on each powered axle as described at paragraph 6.1.2.2. of this GTR. (d) TP1 to include measurement of fuel flow rate

114. ISO 20762 required measurement of intake manifold pressure for verification of engine power by reference to ISO 1585 test conditions. Measurement of fuel flow rate is only required if the confirmation of air fuel ratio according to ISO 1585 is necessary.

115. Experts in the IWG indicated that intake manifold pressure may be insufficient to verify ISO 1585 test conditions especially considering variable atmospheric conditions. Fuel flow rate provides a more precise and additional check.

116. The GTR therefore requires collection of fuel flow rate for TP1 in all cases. To minimize burden, fuel flow rate may be collected from on-board data if its accuracy is shown to the responsible authority. (e) TP1 recommended to measure power input at each inverter if REESS powers multiple inverters

117. ISO 20762 specified that TP1 be performed with measurement of current and voltage at the REESS.

118. The IWG found that this may introduce uncertainties specific to TP1, for electrified powertrains in which the current from the REESS is subsequently routed to more than one propulsion energy converter (i.e. more than one inverter/motor combination) that are deemed likely to experience significantly different electrical conversion efficiencies.

119. For powertrains where the REESS current is routed to more than one propulsion energy converter, this GTR recommends that the input to each inverter be instrumented in addition to the REESS output, unless it is possible to determine net efficiency of the combination, or the efficiencies are the same, as described in paragraph 6.1.3.1. of this GTR. Use of on-board data may be another alternative as allowed in paragraph 6.1.2. (f) Repetition and averaging

120. ISO 20762 does not include a requirement for repetition or averaging of multiple tests. In validation testing, some variation was observed between sequential tests. Korea recommended performing several tests and disregarding the first test result. Subsequent testing confirmed that this practice reduces the variation. The GTR now specifies that five repetitions be conducted and the result be based on an average of the last four repetitions.

121. The GTR also places a limit on the variability of the four averaged measurements, at within ±5 percent of the mean. The variation must be recorded and if it is exceeded, the tests should be performed again, and if the variation cannot be reduced, the result is subject to approval by the responsible authority. (g) Establishment of the "reference point" concept to assure comparable and equivalent results for various HEV architectures

122. The IWG found that the clear identification of reference points for various HEV architectures, and the use of the same reference points for both TP1 and TP2, are important to the expectation that TP1 and TP2 should both deliver a highly similar result. This GTR establishes reference points for common HEV architectures (see Annex 1 of this GTR) and provides a clear definition of "reference point" (see paragraph 3.5.) to assist with the identification of valid reference points for other architectures. (h) Applicability of TP1 or TP2 determined by power flows

123. ISO 20762 did not limit application of TP1 or TP2 to specific powertrain types.

124. The IWG found that the specific flow of power through different electrified powertrain architectures can pose uncertainties for the equitable application of TP1 or TP2 using the specified reference points and measurement points.

125. The GTR therefore includes a set of applicability rules to determine the applicability of TP1 and TP2 based on characteristics of power flow through the powertrain as described in paragraph 6.1.3. of this GTR. (i) Manufacturer to provide hybrid power flow description

126. The IWG found that some electrified powertrains support complex power flows. The specific flow of power that takes place under the maximum power condition is not always clear. This GTR adds a specific requirement for the manufacturer to provide a hybrid power flow description as described in paragraph 6.1.1.1. of this GTR. The description shall also specify recommended measurement points, reference points, and K factor(s), where applicable. The description is intended to provide the authority with concrete information that may be used to determine the applicability of TP1 and TP2 and to assist the authority or third parties with validation and verification. (j) All-wheel drive vehicles to account for each axle independently

127. ISO 20762 did not distinguish between differently powered axles. The GTR adds a specific provision that if a vehicle has two powered axles, each axle shall be tested independently and simultaneously on a 4wd chassis dynamometer or two hub dynamometers, and each may apply a different TP if desired (see text at paragraph 6.1.). (k) Addition of internal validation criteria

128. This GTR introduces a simple mathematical test to reject a result that is inconsistent with the effect of losses in the drivetrain (see paragraph 6.10.). An implied drivetrain efficiency is computed by dividing the power measured at the dynamometer by the sustained power result. Due to drivetrain losses, the quotient should be less than

129. Definitions have been added for several new terms related to system power determination (see paragraph 3.5.). (m) Clarification of gear shifting

130. ISO 20762 did not address the possibility of automatic gear shifting that might occur during the 10-second window of the power test, or the permissibility of manual gear shifting if the gearbox is ordinarily automatically shifted. Text has been added in paragraphs 6.8.6. and 6.9.1. to clarify these issues. (n) Permissibility of validated onboard data for all measurements

131. UN GTR No. 15 allows for the use of on-board data in place of REESS measurements for current and voltage, if the accuracy of the data is demonstrated to the responsible authority. It was noted that such a provision in this GTR could provide an alternative to potentially difficult or impractical instrumentation of inverter inputs or other electrical components under TP1. It was also noted that the added requirement to physically measure the fuel flow rate could be burdensome, and that the use of validated on-board data could also reduce the instrumentation burden for other parameters needed for the power calculation. Text was therefore added in paragraph 6.1.2. of this GTR to generally allow use of on-board data when available, subject to demonstration to the responsible authority that the use of this data meets the accuracy and frequency requirements under paragraph 5.2. (o) Updated equations for calculating system power

132. The equations for calculating system power rating under TP1 and TP2 in paragraph 6.9. have been revised to clarify that the system power rating is the summation of the power calculated at all of the reference points that are applicable to the vehicle powertrain architecture.

133. Both the first and second phases of the validation program provided a wealth of information relating to the practicability and effectiveness of the draft procedure. The opportunity to implement the evolving procedure at several laboratories helped to identify ambiguities in the procedure, as well as evaluate the procedure for the ability to produce an effective characterization of system power in a reliable manner.

134. The differences between the results of TP1 and TP2 that were encountered in the first phase of validation also led to a careful examination of the nature of the problem that the procedure seeks to solve, and the theoretical and physical requirements for a valid solution. This led to the development of the reference point concept, which, when integrated with the procedure, provided (a) a clear technical basis for judging the applicability of TP1 or TP2 to various powertrain architectures, and (b) a strong theoretical basis for the expectation that TP1 and TP2 should yield similar results for powertrains to which both are applicable.

135. Ideally, validation of the procedure would be founded on strong evidence that TP1 and TP2 deliver closely similar results. The latest JRC results from the hub dynamometer tests confirmed a good agreement for a P2 parallel hybrid. However, the validation program was able to produce only limited additional data that would allow a direct comparison between the results of TP1 and TP2 for the same vehicle.

136. One reason is that for some of the powertrain architectures, either TP1 or TP2 is no longer applicable under the revised procedure. These vehicles cannot provide a comparison between TP1 and TP2 because only one is applicable.

137. Another reason is that it was not possible to authentically reproduce all of the aspects of a type approval situation in the validation program. In some cases this limited the ability to perform both TP1 and TP2 in the prescribed manner. For example, in a type approval situation, the manufacturer would have prepared in advance all of the information that is now required for conducting the revised procedure, often relying on proprietary information that was not available to the validation program. This information, such as K factors for TP1 or TP2, hybrid power flow descriptions, and in some cases UN Regulation No. 85 engine test data, were not available, partly because some were new requirements that did not allow the necessary lead time, and partly because of limited motivation for manufacturers of the selected vehicles to provide this proprietary information.

138. Aside from limited opportunity to directly compare TP1 and TP2, there are several persuasive reasons to have good confidence in the ability of the revised procedure to deliver valid results.

139. In the revised procedure, it is now assured that both TP1 and TP2 measure power at the same reference points. This eliminates a cause of some previously identified differences, which were related to the implicit use of sometimes inconsistent reference points in the earlier version of the procedure.

140. The revised procedure also makes it clear whether or not a given TP is applicable to a given vehicle, preventing the possibility of applying a TP for which the powertrain architecture cannot support its use, and leading to the delivery of only a single result.

141. Additionally, the validation program provided additional evidence that the maximum power of the vehicles tested can be reliably commanded by the fixed-speed dynamometer method.

142. The primary remaining potential source of error between the two TPs would be measurement error. Requirements for measurement accuracy and frequency are clearly identified in the procedure, and align with similar requirements in ISO 20762 and UN GTR No.

143. At this time, this GTR specifies a reference method but not a candidate method. A candidate method, which would not require dynamometer testing but instead would be based on the results of component tests, would potentially allow a vehicle power rating to be determined at a lower expense. Future development and validation of a candidate method remains a possibility for future work. D. Technical feasibility, anticipated costs and benefits

144. The specification of a test procedure for power determination will remove significant uncertainty that manufacturers now face in communicating the power level of electrified vehicles both to the public and to regulating authorities, and resolves the question of how to determine a system power rating for electrified vehicles for use with WLTP.

145. Initially the adoption of the procedure may bear some costs for vehicle manufacturers, technical services and authorities, at least considered on a local scale, since some test equipment and procedures may have to be upgraded. However, these costs should be limited since such upgrades are done regularly as adaptations to technical progress. Related costs II. 1. 2. 2.1.

2.1. This UN GTR applies to vehicles meeting all of (a)(b)(c): (a) hybrid electric vehicles, or pure electric vehicles with more than one propulsion energy converter; (b) category N1, or category M with technically permissible maximum laden mass not exceeding 3,500 kg; (c) if a hybrid electric vehicle, at least one electric machine contributes to propulsion under the maximum power condition.

2.3. When determined according to this UN GTR, the resulting vehicle system power rating may be considered comparable to the power rating traditionally assigned to conventional vehicles, which is the power rating of the internal combustion engine.

2.4. The following documents are referred to in the text in such a way that some or all of their content constitutes requirements of this document, notably ISO 1585 and UN Regulation No. 85 (engine power), and UN GTR No. 15 (WLTP).

3.1.1. "Technically permissible maximum laden mass" means the maximum mass allocated to a vehicle on the basis of its construction features and its design performances.

3.1.2. "Fixed speed mode" means the operating mode of the dynamometer in which the dynamometer absorbs the power output of the vehicle so as to maintain the vehicle at a fixed dynamometer speed.

3.2.1. "Powertrain" means the total combination in a vehicle of propulsion energy storage system(s), propulsion energy converter(s) and drivetrain(s) providing the mechanical energy at the wheels, plus peripheral devices.

3.2.2. "Peripheral devices" means any energy consuming, converting, storing or supplying devices where the energy is not used for propulsion but which are essential to the operation of the powertrain and are therefore part of the powertrain.

3.2.3. "Auxiliary devices" means energy consuming, converting, storing or supplying non-peripheral devices installed for purposes other than propulsion and therefore not part of the powertrain.

3.3.1. "Energy converter" means a system where the form of energy output is different from the form of energy input.

3.3.2. "Propulsion energy converter" means an energy converter of the powertrain, not a peripheral device, whose output energy is used directly or indirectly for propulsion.

3.3.3. "Charge-depleting operating condition" means an operating condition in which the energy stored in the REESS decreases on average until transition to charge-sustaining operation.

3.3.4. "Charge-sustaining operating condition" means an operating condition in which the energy stored in the REESS is, on average, maintained at a neutral charging balance level.

3.3.5. "Category of propulsion energy converter" means (i) an internal combustion engine, (ii) an electric machine, or (iii) a fuel cell.

3.3.6. "Energy storage system" means a system which stores energy and releases it in the same form as was input.

3.3.7. "Propulsion energy storage system" means an energy storage system of the powertrain, not a peripheral device, whose output energy is used for propulsion.

3.3.8. "Category of propulsion energy storage system" means (i) a fuel storage system, (ii) a rechargeable electric energy storage system, or (iii) a rechargeable mechanical energy storage system.

3.3.9. "Form of energy" means (i) electrical energy, (ii) mechanical energy, or (iii) chemical energy (including fuels).

3.3.10. "Fuel storage system" means a propulsion energy storage system that stores chemical energy as liquid or gaseous fuel.

3.3.11. "Electric machine" means an energy converter transforming between electrical and mechanical energy.

3.3.12. "Off-vehicle charging hybrid electric vehicle" (OVC-HEV) means a hybrid electric vehicle that can be charged from an external source.

3.3.14. "Hybrid vehicle" means a vehicle with a powertrain containing at least two different categories of propulsion energy converters and at least two different categories of propulsion energy storage systems.

3.3.15. "Hybrid electric vehicle" means a hybrid vehicle with at least one electric motor or motor-generator and at least one internal combustion engine as propulsion energy converters.

3.3.16. "Pure electric vehicle" (PEV) means a vehicle with exclusively electric machines as propulsion energy converters and exclusively rechargeable electric energy storage systems.

3.3.17. "Rechargeable electric energy storage system" (REESS) means a propulsion energy storage system that stores electrical energy and is rechargeable. A battery whose primary use is starting, lighting or other auxiliaries is not a REESS.

3.4.1. "Driver-selectable mode" means a distinct driver-selectable condition which could affect emissions, fuel and/or energy consumption, or maximum system power output.

3.5.1. "Test procedure 1" (TP1) means a test procedure for determining a vehicle system power rating via measured electrical power and determined ICE power.

3.5.2. "Test procedure 2" (TP2) means a test procedure for determining a vehicle system power rating via measured torque and speed at the axles or wheel hubs.

3.5.3. "Power determination reference point" means a point in the mechanical power flow path of a powertrain where any portion of the mechanical energy that drives the wheels under the maximum power condition is first produced as mechanical energy by a propulsion energy converter from a propulsion energy storage system.

3.5.5. "Speed of maximum power" means the fixed speed setting of the dynamometer at which a maximum accelerator pedal command, given for at least ten seconds in power-rating mode, delivers the greatest peak power to the dynamometer.

3.5.6. "Maximum power condition" means the condition in which the vehicle is on a dynamometer in power-rating mode, the dynamometer is in fixed speed mode at the speed of maximum power, and the maximum accelerator pedal command is given for at least ten seconds.

3.5.7. "Vehicle system power rating" means the total power transmitted through all of the power determination reference point(s) as determined by TP1 or TP2.

3.5.8. "Mechanical energy path" means a distinct parallel path within a drivetrain that conducts a portion of the total mechanical energy passing through the drivetrain.

5.1.1. The power absorption capacity of the dynamometer in fixed speed control mode shall be sufficient for the maximum power of the vehicle. Due to the short duration of maximum power (about 10 seconds), a short-duration power rating of the dynamometer may apply with approval of the Type-Approval Authority.

5.1.3. Cooling fan: a current of air of variable speed shall be blown towards the vehicle sufficient to maintain proper system operating temperatures; above 5 km/h the blower outlet air velocity shall equal the dynamometer speed within plus or minus 10 percent. Excessive cooling is prohibited.

5.1.4. Soak area: the soak area shall have a temperature set point of 25 degrees C, with the actual value within the specified tolerance; at the request of the manufacturer 25 degrees C may be replaced by 23 degrees C.

5.2.1. Measurement items and accuracy: measurement devices shall be of certified accuracy as shown in Table 2, traceable to an approved regional or international standard (e.g. voltage and current plus or minus 0.3% FSD or 1% of reading; axle torque plus or minus 6 Nm or 0.5%).

5.2.2. Measurement frequency: all items in Table 2, unless otherwise specified, shall be measured and recorded at a frequency of at least 10 Hz; atmospheric pressure and room temperature are recorded at the start and end of vehicle operation.

6.1.1.1. Power flow description: the manufacturer shall provide a power flow description sufficient to identify the energy flow paths and conversions during the maximum power condition, from each propulsion energy storage system to each powered axle, including REESS-powered DC/DC and high-voltage auxiliaries.

6.1.1.2. Energy conversion factors (K factors): where TP1 is performed the manufacturer shall provide the electrical energy conversion efficiency K1 between each electrical measurement point and the corresponding reference point; where TP2 is performed, the mechanical conversion efficiency K2. The provided values are subject to verification by the responsible authority.

6.1.1.4. Other information: the manufacturer shall specify the normal operating range for each operational metric listed in paragraph 6.8.1.

6.1.2.1. TP1-specific measurements: TP1 additionally requires measurement of voltage and current at the REESS or inverter inputs, engine speed, intake manifold pressure, and fuel flow rate.

6.1.2.2. TP2-specific measurements: TP2 additionally requires measurement of torque and rotational speed at the powered axle(s) or wheel hubs.

6.1.3.1. Applicability of TP1: TP1 applicability requires that the power passing through all reference points can be accurately determined by performing the prescribed procedure.

6.1.3.1.1. Where the power flow description indicates that the current from each REESS powers a single electric machine, current and voltage at each REESS output can be determined, and an accurate K1 factor is provided, TP1 is applicable. Power at R [kW] = (U x I / 1000) x K1.

6.1.3.2. Applicability of TP2: TP2 applicability requires that the power passing through all reference points can be accurately determined; each powered axle is evaluated separately, and TP2 is applicable only if applicable to all powered axles.

6.1.3.2.1. Where torque to the axle originates from a single reference point and an accurate K2 factor is provided, TP2 is applicable to that axle. Power at R [kW] = (2*pi x torque x rev-per-second / 1000) / K2.

6.1.3.2.2. Where the axle torque is a combined torque contributed from a set of reference points routed via the same mechanical path, and an accurate combined K2 is provided, TP2 computes the sum of the power at that set of reference points.

6.2.1. Rollers (chassis dynamometer only): the chassis dynamometer rollers shall be clean, dry and free of foreign material that could cause slippage.

6.2.2. Tire slip (chassis dynamometer only): measures may be taken to stabilize tire slippage that may occur under maximum power, and shall be recorded.

6.2.3. Dynamometer warm-up: the dynamometer shall be warmed up per the manufacturer recommendation or other suitable method until internal friction losses stabilize.

6.2.4. Dynamometer control: during vehicle conditioning the dynamometer may be operated in road-load mode.

6.8.1. General: the test shall be carried out in accordance with paragraphs 6.8.3 to 6.8.8 and 6.9 to 6.10. The test shall be stopped immediately if a powertrain warning indicator turns on. Operational metrics shall be monitored and recorded throughout.

6.8.3. Vehicle conditioning: operate the vehicle until the manufacturer-specified normal operating temperature ranges are reached and stabilized. Initial conditioning is performed in power-rating mode at 60 km/h at the vehicle road load for at least 20 minutes, or as recommended by the manufacturer.

6.8.5. Vehicle operation: for vehicles with driver-selectable modes, select the mode for which a vehicle system power rating is desired and record it as the power-rating mode; place the dynamometer in fixed speed mode set to the speed of maximum power and allow it to stabilize.

6.8.6. Power test: the maximum accelerator pedal command shall be given by pedal position or vehicle communication network for at least 10 seconds, as rapidly as possible; the gear shall be selected as recommended by the manufacturer for a typical driver to achieve maximum power.

6.8.7. Repetition of the power test: the power test shall be repeated for a total of five repetitions; prior to the second and subsequent repetitions the REESS shall be adjusted.

6.8.8. End of vehicle running: at the end of vehicle running the operational metrics shall be recorded, and the vehicle and measurement devices shall be stopped.

6.9.1. General: for each of the 2nd, 3rd, 4th and 5th repetitions, the time-series data shall be analysed to calculate peak (3.5.9) and sustained (3.5.10) vehicle system power; the rating is the mean of the four repetitions, and each individual value should be within plus or minus 5% of the mean.

6.9.2.1. For reference points consisting of ICE power: first determine the ICE power from the full-load power curve as a function of engine speed, subject to confirmation of intake manifold pressure and fuel flow rate (use the curve value if both are within 5% of the certification values; otherwise re-test under the observed conditions).

6.9.2.2. For reference points consisting of electric machine power where the measurement point is the REESS output: Ri [kW] = (U_REESS x I_REESS / 1000 - P_DCDC - P_aux) x K1, where P_DCDC is power to the DC/DC converter for 12 V auxiliaries (1.0 kW or measured) and P_aux is power to high-voltage auxiliaries other than P_DCDC.

6.9.2.3. For reference points consisting of electric machine power where the measurement point is the inverter input: Ri [kW] = (U_input x I_input / 1000) x K1.

6.9.3.1. Calculation for TP2: the vehicle system power is the sum of the power at each reference point. Ri = P_axle / K2, where P_axle [kW] = (2 x pi x axle speed[rev/s] x axle torque[Nm]) / 1000.

6.9.3.3. Corrected vehicle system power rating for TP2: where TP2 does not deliver a distinct ICE power portion (e.g. power-split), derive PICE by subtracting the non-ICE reference points (measured by TP1) from the summed power, correct PICE per ISO 1585:2020 paragraph 6, and recompute the system power as the corrected ICE power plus the non-ICE reference point power.

1.1. Both TP1 and TP2 transform a specified set of vehicle test measurements into a system power rating referenced to one or more power determination reference points.

1.2. The power determination reference point is the point in the electrified powertrain closest to the engine output shaft of a conventional vehicle.

1.3. The power determination reference point is defined in paragraph 3.5 as the point in the mechanical power flow path where the mechanical energy that drives the wheels is first produced.

1.4. For complex electrified powertrains, the reference points may depend on the given operating mode or the given power rating sought for the vehicle.

2.1.1. The power determination reference points capture the total mechanical power delivered to the road under the maximum power condition.

2.2.1. For a parallel architecture (example in Figure 25), the reference points are generally (a) R1 at the engine and (b) R2 at the electric machine.

2.2.2. In Figure 25 the electric machine EM drives the engine output shaft directly; the reference points are R1 and R2.

2.3.1. A power-split architecture (example in Figure 26) often uses one or more planetary gear sets; the reference points are R1 and R2REESS.

2.3.2. Here TP1 determines R1 from engine speed, manifold pressure and fuel flow (full-load power curve) and R2REESS from the REESS measurement with K1. (TP2 is not applicable.)

2.4.1. A pure series architecture (example in Figure 27) has no mechanical connection between the engine and the road; the reference points are R1 and R2REESS.

2.4.2. TP1 is performed as in 2.3.2. (TP2 is not applicable to a pure series architecture.)

2.5.1. Where two or more axles drive the vehicle under the maximum power condition, both axles are tested simultaneously.

2.5.2. Here TP1 determines R1 from engine measurements and the electric-machine reference points from the inverter inputs with the applicable K1 factors.

2.6.1. For other architectures or variants not listed in this annex, the reference points and the applicability of TP1 or TP2 shall be determined by the manufacturer and the responsible authority.

1. The speed of maximum power (defined in paragraph 3.5) is the speed at the maximum of the power-speed relationship (see Figure 29).

2. The speed of maximum power is determined by the manufacturer or the responsible authority by the procedure described in this annex.

3. The speed of maximum power is identified by operating the vehicle at a series of fixed dynamometer speeds (operating points) to find the speed at which power is greatest.

4. The series of operating points shall be spaced closely enough to identify the speed of maximum power with sufficient confidence.

5. At each operating point, the power delivered to the dynamometer is determined from dynamometer power data when available, or from measured torque and speed.

6. The determined speed of maximum power is reported as an integer in kilometres per hour (km/h).

7. Where the vehicle manufacturer has specified the speed of maximum power and verification is desired, test at points above and below the specified speed.