Vehicle Testing in ENEA Drive-train Test Facility

Transcription

1 Vehicle Testing in ENEA Drive-train Test Facility Giulia LO BIANCO, Giovanni PEDE, Angelo PUCCETTI, Ennio ROSSI ENEA- Energy Department Giorgio MANTOVANI Copyright 2000 Society of Automotive Engineers, Inc ALTRA ABSTRACT Among the research and development subjects of ENEA (Italian National Agency for New Technology, Energy and Environment) an important theme is the study of innovative vehicles with high energy efficiency and low emissions. ENEA has set up an infrastructure in order to execute research and development activity for hybrid and electrical vehicles. The paper deals with features and working methods throughout the testing campaign on the ALTRA minibus drive-train, a 6m. series hybrid developed by IVECO and ANSALDO. The test demonstrated that the energy consumption is strong correlated to the power management strategy. Subsequently, a testing work was carried on with a controller developed by ENEA and University of Pisa, by using the roll bench of ENEA, while the IVECO tested on the road an other control system. All these tests convalidates the need of a controller to achieve fuel economy. INTRODUCTION ENEA (Italian National Agency for New Technology, Energy and the Environment) is conducting a large Research and Development program on innovative vehicles with high energy efficiency and low environmental impact. This program is partly supported by the Italian Ministry of Industry with the aim of achieving significant improvements of technology and includes the participation of research institutions and industry. Testing and evaluation activities play a strong role in this program. Components and subsystems are fully characterized using a Battery Test Facility and a new Drivetrain Test Facility, while complete vehicles are tested on a Roller Bench Test Facility and in a dedicated circuit or on regular roads. The main characteristics of the conventional facilities have been thoroughly described previously [1, 2 ] and this paper will show the main characteristic of the Drivetrain Testing Facility. In the last year, a number of experiences, based on the testing of a hybrid drive system, have been performed. The rationale and the first results of the testing are as follows. POWER MANAGEMENT IN A SERIES HYBRID-ELECTRIC VEHICLE In a series hybrid vehicle, as we have two different power units (an electric generator and an energy storage system), at any time in driving (except in the braking mode) the sum of the generator power, Pg, and the battery power Pb, multiplied by the electric motor efficiency (? m < 1), should be equal to the vehicle load power PL, as expressed by: 1) (Pg (t)+ Pb(t) ) x? m = PL(t) During the development phase, to decide about the sizing of the generator set and the storage system according to Equation 1), the maximum possible value for PL, should be considered, related to the desired vehicle performance. In this phase we can choose among different values of the ratio Pg/Pb, as to design different types of vehicle, starting from range-extender hybrid, where Pg is minimal and all the power is provided with the battery, recharged during stops and at night, up to the so called dieselelectric hybrid, where Pb is zero and Pg is equal to the vehicle load power, plus the electrical losses in the transmission, control system and traction motor. For a general purpose hybrid, the generator power output should match the vehicle load power PL at the desired maximum cruising speed. 1

2 Pg and Pb are the maximum power of the generator and of the battery system; in driving, when the vehicle load power PL(t) varies from zero to the maximum, equation 1) holds true and the control system of the drivetrain has to determine, second by second, the power output of each power unit. Minimizing fuel consumption and exhaust emissions can be done in many ways, provided the total power output meets the instantaneous requirements. For an estimate of an HEV energy use under a variety of driving cycles, computer simulation models are of paramount importance. The results of the simulation can be used both during the development phase of the vehicle, as an important feedback for a proper selection of the energy capacity of the battery pack, (oversized battery pack would cause vehicle overload and undersized pack can not supply adequate power for the needs of the vehicle), and in driving, for a proper management of the drive system. Essentially, two types of control strategies are possible for the generator 1) the load following, where Pg(t) follows the vehicle power demand with the maximum transient compatible with exhaust emission control, 2) the on-off (duty-cycle) strategy, where the power generator set is chosen according to the estimated average requested power and the adjustments are made by setting on and off the internal combustion engine. The first allows to reduce the battery pack, by increasing, second by second, power supply from the generator according to the needs of the vehicle. The second one is simpler and allows to tune the engine to its optimal operating point, but it needs to know the average medium power. As for the battery, it is important to achieve the maximum efficiency both in charging and in discharging, by keeping the SOC within a specific range, which depends on the battery. In fact, battery efficiency is related to internal resistance, that expresses the combined effects of electrical resistance and chemical behavior, and this parameter tends to rise at the end of the charge and discharge, that is for low and high values of the SOC. So, there is an optimal range throughout SOC where the battery has an optimal charging and discharging performance. The testing campaign on ALTRA minibus drivetrain, a 6m. series hybrid developed by IVECO and ANSALDO was conceived to experimentally demonstrate these concepts, and to find the optimum values for generator power level and battery State-of-Charge. ENEA DRIVETRAIN TESTING FACILITY The Drivetrain Testing Facility [3] is aimed at experimenting and fully characterizing complete drivetrains (and subsystems) for pure electric and hybrid vehicles. The concept around which it has been planned is the following: the hybrid vehicles can be designed according to various architectures, but they all have on board a power generator, of various types, an energy storage system to back up the power generator, one or more electric motors, an apparatus for control and conversion of electric energy. So the testing infrastructure is composed of four interconnected experimental areas, the first dedicated to generation of electricity by means of motorgenerators, fuel cells, turbogenerators; the second for energy storage testing, mainly suited for batteries but also suitable for flywheels or supercapacitors; the third and the fourth are equipped with engine/motor test benches in order to simulate the vehicle load, Fig. 1. Inverters and control equipment are monitored by dedicated instrumentation. Each section acts as an experimental island, able to operate alone or jointly with the other sections. In fact, all the sections are remotely controlled and managed in order to create a different assembly of the drivetrain. The facility has been sized to allow the testing of drivetrains and subsystems of small and medium-sized vehicles (up to minibuses), but an extension to larger vehicles is possible. Figure 1 presents the layout of the facility with the main sections and equipment. POWER GENERATION SECTION. This section allows the characterization and test of APU (Auxiliary Power Unit) subsystems for onboard use in electric and hybrid vehicles: thermal generators and fuel cells. Different thermal engines can be used: diesel engine, gasoline engine, gas turbine, hydrogen engine. The section permits the operation of thermal generators up to 30 kva and fuel cells up to 20 kw (peak power) and includes:?? liquid and gaseous fuel supply (diesel, gasoline, methanol, hydrogen, methane);?? electric generation up to a maximum of 500 A at a voltage 150 V;?? displacement of exhaust gases;?? cooling systems for combustion engines;?? data acquisition system with a set of sensors for measuring main parameters (partial and global values). 2

3 ENERGY STORAGE AND MANAGEMENT. This section has bi-directional AC/DC converters with many duties: charge and discharge of batteries at a defined depth of discharge (DOD); power traction electric motors in case the battery is not used; connection of the APU (or the driving motors operating as generators) to the grid, acting as a DC load with programmable functions (constant current, constant power and constant resistance). The section can be operated between 60 V and 360 V up to a maximum power of 100 kw. A large climatic room (2 m3) is part of this section: batteries can be thermally controlled and also tested with a dedicated battery cycler. The data acquisition system allows the measurement of the main parameters of the battery under test. DRIVING MOTORS. This section is composed of two test rooms for electric motors and thermal engines. The rooms are equipped with circuits for the displacement of exhausted gases and for motor cooling.?? Motor Test Room 1. ; Fig.2 There are two 30 kw test benches of the type reversible electric machine operating on 4 quadrants.?? Motor Test Room 2., Fig. 2. There is a 100 kw test bench of the type reversible electric machine operating on 4 quadrants. The test benches are completely programmable in order to simulate different road loads. TESTS PROGRAM. The complete Drivetrain Test Facility is able to execute two categories of tests: the integrated drivetrain and the subsystems. Principal driving patterns and standard duty cycles (ECE-15, SAE, SFUDS, DST, and so on) can be programmed and automatically performed in any section of the facility. In the integrated drivetrain test the test facility simulates the driving patterns of any vehicle, composed of the drivetrain (electric motor and APU) and the battery, on a defined road profile. A maximum of 200 phases can form the speed cycle. It is also possible to consider the variation of the vehicle mass during stops (variations of payload) and include the energy recovered during braking. The entire test runs are centrally programmed and controlled from a remote control room. The data acquisition system issues many preelaborated outputs:?? speed versus time?? specific energy consumption at various speeds (with varying load and slope)?? global energy consumption in function of the distance traveled?? state of charge of the battery?? distance traveled per test run. THE ALTROBUS The ALTROBUS hybrid bus ( a series concept hybrid bus designed accordingly to the following general lay-out ) was developed in cooperation with ANSALDO RICERCHE e FIAT IVECO and realized by the ALTRA Society [4]. Fuel ICE AC Generator Generation System P=K Battery Charger Battery Pack Electric Propulsion System Controller DC/DC Chopper Brake Accelerator DC Motor Gear Wheel Wheel There are two versions: the first one, the 490 type, has a weight of 12.8 ton without load and can carry about 85 passengers; the second one, named Daily, Fig. 3, is 6 m long, has a net weight of about 4.2 ton and can carry 21 passengers. The 490 bus has a 2.8 diesel engine, an electric motor of 164 kw maximum power and lead acid batteries with 60 kwh capacity. The hybrid bus range is km in the thermal mode and 20 km in the electric mode. The vehicle cost is about twice the cost of a conventional bus, but it has a significant reduction in pollutant emission (83% Nox, 60% as CO, 80% as HC) and noise. The Daily has a 1.2 l diesel engine, an electric motor of 33 kw maximum power and lead acid batteries with a capacity of 20 kwh. The vehicle range is km in the thermal mode and 20 km in the electric mode. Maximum speed is 60 km/h and the emission reduction is similar to the one of the 490. Other main characteristics for the two vehicles are the following:?? the thermal engine is compact and is programmed to work at the point of maximum efficiency?? the generator power can be varied only out of line, choosing among a set of 4 values (6,7,8,9 kw for the Daily). During the test, the power moves around the set point, accordingly to the AC/DC Converter curve :?? the power Conditioner (AC/DC Converter) works with a constant power/variable voltage curve (hyperbolic path in the diagram voltage/current) so the battery automatically feeds the motor, when current is high and voltage is low, on the contrary is recharged when current is low and voltage is high These two models have already been used in many cities, with a cumulative range of more 3

4 than 2 million kilometers. Up to now, the total fleet is composed of 91 busses, of which 43 belong to the 490 type anthe remaining 48 to the the Daily one. Fig. 4 shows the Altra Daily drive train characteristics. ALTRA DRIVETRAIN TESTING Testing of the drive train [ 5 ] occurred by installing the major components, engine/generator, battery pack, electric motor and electronic battery charger in section 1, 3, 4 of the Drivetrain Test Facility, Fig. 1. The 100 kw test bench applied load to the traction motor duplicating loads experienced during normal driving conditions, the data acquisition system measured and recorded the main parameters of the components under test. A separate test was performed on the battery pack, aimed at obtaining the battery efficiency as a function of the battery State-of-Charge. TEST PROCEDURES. The aim of the test was to measure the specific diesel fuel consumption of a hybrid vehicle during the ECE 15 cycle, as a function of the battery S.O.C. (State Of Charge) or the D.O.D. ( Depth of Discharge, D.O.D., = 1 - S.O.C. ) and of the generator power level. Accordingly, during the test the battery charger power changed in the range 6? 9 kw, and, for each power level, the test was performed with 3 different battery D.O.D. 10%, 40% and 90%, for a total of 12 tests: D.O.D. = 10% 40% 90% 6 kw n. 1 n. 5 n. 9 7 kw n. 2 n. 6 n kw n. 3 n. 7 n kw n. 4 n. 8 n. 12 The D.O.D. was set by monitoring and adjusting, with cycles all electric or idle periods, the battery open circuit voltage, O.C.V. accordingly to a standard internal procedure. A specific test was performed to verify the accordance between the open circuit voltage measured after a rest period of 3 hours, according to the standard, and the same measure after 20. The difference between the two values ( and the linearity between D.O.D. and O.C.V.) remained the same up to D.O.D. 70%, so the values measured after the shorter rest period was considered suitable to give useful indications for the D.O.D. of the battery. Each test was performed by repeating a minimum of 8 basic urban cycles (total time test:1568 s). When necessary to adjust (reduce to the starting value) the battery DOD, more urban cycle with the generator off were performed. During the test the diesel fuel consumption was measured and the system voltage and the three current values were recorded (dc motor, battery system, battery charger). V Open Circuit Voltage D.O.D. 180 The test to estimate the battery efficiency was performed by subsequently discharging and charging the battery, then measuring the ratio between charged and discharged energy; in this way we measured the global charging and discharging efficiency, given by the expression:? disch x? ch =? glob The discharge current was C/5, ( I = 20 A), the charging was performed accordingly the manufacturer s standard IU (maximum voltage 230 V). Partial charging and discharging values are calculated by differencies between two test in a row, partial efficiencies by the ratio between the so obtained values. TEST RESULTS. During a typical urban drive cycle (we used the urban part of the European omologation driving cycle, NEDC), the current profiles of motor, generator and batteries follow the underlying trends, accordingly to the balance-of-power relation 1) : 20 4

5 Currents (A) Time (s) Battery Motor Generator Thanks to the energy storage system, the generator power ranges around a mean value, with transients much lower than in a conventional vehicle, so that emissions are consequently lower by one order of magnitude [8]. 10 9,5 9 8,5 8 7,5 7 6,5 6 5,5 Generator Power (kw) Time (s) In fact, extra power requirement is supplied by the battery, whose voltage ranges from a minimum and a maximum value, set by the manufacturer, as shown in the picture below : consumption. The last 3 tests, related to the DOD 90%, are not reported, because the battery power was not sufficient to perform the last part, up to 50 km/h of the urban cycle, so the fuel consumption of these groups of tests is not comparable to with the others. The specific consumption ranges from 16.6 to 21.7 l/100km, and the best values were obtained at 8 kw. The D.O.D. optimum value is around 60%, upper values show poor fuel economy, lower values are insufficient to drive the vehicle accordingly to the ECE 15. The generator efficiencies as a function of the power setting is shown in the next figure, that partially explains the fact that the generator has quite the same consumption (per minute) at different power setting. g/kwh Specific Compsuption kw 2400 r.p.m. As regards battery efficiencies as a function of the SOC, the test, Fig. 6, gave an equivalent result, showing a maximum in the region between 30% and 70%: Charging and Discharging Efficiency 240 Battery Voltage (V) Time (s) % 50% 100% State of Charge Efficiency When these values are reached, the generator shifts to a slightly higher or lower power. Fig. 5 shows the main results of the trials for each test (8 or more urban cycle). For each test, there are 3 different ways ( hybrid, only electic, and recharging at the stops) to work on the drive system, and the related specific consumption are indicated in l/100 km or l/min for the recharging mode during the stops. The total consumption is divided by the total distance travelled, thus obtaining the specific TEST RESULTS DISCUSSION. The test demonstrated that the energy consumption is strong correlated to the power management strategy, and to achieve the best results we have to satisfy two conditions: 1. operate as close as possible to the optimal operation points of generator and battery, characterized by laboratory tests. 2. set the generator close to the average power of the cycle, so to reduce the 5

6 energy dissipation in the storage system without, on the other hand, depleting SOC under a minimal value, so reducing the available peak-power to an unacceptable one. According to the test results, the best value for the generator power setting was around 8 kw, that is close to the average power (10 kw) of the cycle, (with no braking energy recovery). Actually this is not the optimum value to achieve the generator minimum specific compsuption, as for this engine this point is around 12 kw, but the chosen range of values (6? 9 kw) was a compromise between emissions (smoke) and economy requirements. In this case, the adoption of an indirect injection diesel engine was a handicap for the hybrid version of this vehicle, compared to the conventional version equipped with a direct injection diesel. The battery efficiency is good enough for a fairly wide range, so a sophisticated S.O.C. depending control system is probably not necessary, provided that a very high S.O.C. (90%) is avoided. On the contrary, it is of the utmost importance do not recharge during stops (particularly at low power) or perform all electric cycle, so an automatic generator control based on the periodic measure of the average requested power, for instance every 5? 10 minutes, is highly recommended. ROLLER BENCH EXPERIENCES In the framework of the collaboration between University of Pisa and ENEA, it has been decided to make roller bench tests on the Altrobus, to evaluate the real effectiveness of different control algorithms in reducing fuel consumption of electric hybrid vehicles, industry, with a controller specifically developed by the University of Pisa [ 6, 7]. The power of today s microcontrollers is such that low cost (10-20 US$) commercial microcontrollers comply with this requirement. In particular the microcontroller chosen has a 16 bit CPU, a 12 bit A/D conversion unit, and 32 kb of RAM. The microcontroller starts-up and shuts down the engine and controls the operational speed so that the generator could work at its best point and the energy stored in the battery remains within pre-defined limits E min -E max. The output power of the DC source is maintained constant. Using this controller, the effects of the choice of different valued of E min and E max can be experimentally evaluated on the vehicle consumption. The testing campaign was performed according the european and american urban driving cycles. The UDDS is a cycle severe enough to force the hybrid to a charge depleting mode, instead of the conventional charge sustaining mode, so to investigate also this part of the working range. During each test run a set of parameters has been measured, recorded and elaborated: 1. Vehicle speed 2. Battery current 3. Electric motor current 4. Generator current 5. Battery voltage 6. Fuel consumption 7. SOC 8. Travelled distance 9. Energy for recharging the battery from mains 10. Test time Hereafter the test results for the european and the american cycle are summarized in an homogenous way. The figures below show the battery SOC during test. % % S.O.C seconds ECE 15 URBAN S.O.C seconds UDDS COLD STABILIZED CYCLE Main parameters Units NEDC UDDS Driven distance Km

7 Total time Hours Average speed km/h ? SOC % Fuel consumption Total cons. Ah Wh kg/h kg Wh energy Wh Wh/km l/100 km Comparing the fuel consumption during the NEDC with the best case for the vehicle without controller, Fig. 5, it is apparent the obtained improvement, so the need for such a device on a hybrid vehicle is confirmed. APPLICATION EXPERIENCES AND INDUSTRIES TREND ON HYBRID VEHICLES MANAGEMENT SYSTEMS Also the experiences on hybrid buses fleet use underline the need of a correct control of the generation function, to optimize the fuel consumption and to reach the best system performance. For this reason Altra developed a first generation of energy supervisor system able to manage the working time and the stop time of generator [8]. This device, applied on vehicles in real service, show consumption decreasing up to 30%. In order to increase the performance flexibility of the vehicles, according to the technical development of the components and hybrid protection systems, Altra is developing a new release of energy supervisor system with the following features:?? Digital interface (CAN BUS) with vehicle and components;?? ICE speed and power control;?? Over temperature and over discharging batteries protection;?? Increased control logic;?? Battery temperature feed on charger control;?? Increased battery life;?? Increased system performances;?? Decreased 5% more consumption;?? Increased battery reliability. The new generation control logic of Altra supervisor is based on tree main controls:?? Discharging battery performance in electric mode;?? Ratio between average working power and set point power of the batteries charger in hybrid mode;?? Batteries state of charge. CONCLUSIONS To minimise the fuel consumption of a series hybrid electric vehicle there is the need of an automatic controller that : 1. operates as close as possible to the optimal operation points of generator and battery, characterized by laboratory tests. 2. keeps the ICE rotational speed adapted to the power to be delivered to the DC busbar 3. keeps into account the energy level within the battery; therefore an accurate state-ofcharge estimating algorithm has to be included in the controller 4. finally, the energy levels have to remain within a domain defined by the maximum battery deriverable power and the need to avoid gassing during charge. In the next months, research centre, university and industry will work in the same project in order to integrate different experiences and different approach with the aim to optimise the control system performances. REFERENCES 1. Battery and EVs Field Testing at ENEA Research Centre, Conte M., Pede G., Vellone R., EVS-13, Osaka, October A Close Up View of a Comparative Test on Electric Vehicle Batteries in ENEA, Pede G., De Andreis L., Indiano E., E., Romano P., Vellone R., SAE - Future Transportation Technology Conference -, San Diego, August, On the bench, G. Bernardini, M.Conte, L.De Andreis, G.Pede, E.Rossi & R. Vellone, Electic and Hybrid Vehicles Technology Hybrid Vehicles Projects and realizations in Italy, Mattucci, G.Pede, R. Vellone, E.Durelli, D. Mesiti, Workshop INRETS, Lione, May Studio sperimentale di un sistema di trazione ibrido e di un veicolo ibrido, E.Rossi, A.Puccetti, RTI Enea, Rome, October Experiences on control of internal combustion engines of series hybrid vehicles, G.Pede, M.Ceraolo, R. Giglioli, EVS-16, Pechino, Control of series Hybrid Electric Vehicles : alghorithms and experimental test, M.Ceraolo, E.Rossi, G.Pede, EVS-17, Montreal, Hybrid bus procedures for compsumption and pollution evaluation, A.Barbieri, 7

8 E.Durelli, G.Mantovani, ALTRA S.p.A. Geneva; A. Pini Prato, University of Genoa, EVS 15, Brussels, October

9 ) Twin 30 kw test benches; 2) 100 kw test bench; 3) Power unit/fuel cell; 4) Energy storage and management; 5) Control Room; 6) Assembly Area Figure 1. Layout of the ENEA Drivetrain Test Facility. Figure 2. View of the Motor Test Room 2 with the 100 kw test bench. 9

10 Fig. 3 Altra Daily hybrid bus GENERATION SYSTEM ELECTRIC PROPULSION DRIVE SYSTEM Internal Combustion Engine Separately Excited DC Motor Fuel Diesel oil Rating Power (@ range speed rpm) 22 kw Displacement 1204 cc Rating Armature Voltage 192 V Maximum Power (@3600 rpm) 24 kw Rating Field Voltage 85 V Rating speed 2200 rpm Maximum Power 32.5 kw (@ 2080 rpm 60 sec overload) Regulation mechanica l Maximum speed 4800 rpm Synchronous PM Machine Maximum Torque (@ 1200 rpm) 172 Nm Rating Power (@ 2200 rpm) 10 kw dc-dc Electronic Power Converter Rating Voltage 220 V Current-controlled Mode Separate Control of Armature and Field current ELECTRONIC BATTERY CHARGER Regenerative Breaking Rating Power 10 kw Rating Voltage 192 V MECHANICAL TRASMISSION SYSTEM Fixed Gear Ratio 1:2.260 STORAGE SYSTEM Lead Acid Battery Cell 96 Capacity(C 5 ) 100 Ah Fig. 4 Altra Daily Drive Train Characteristics 10

11 test n kw In hybrid mode Recharging during stop Only Electric Total consumption km fuel, kg l/100 km Minutes fuel, kg l/min km km fuel, kg l/100 km D.O.D ,5 1,1 15, ,42 0,051 0,00 8,50 1,52 21,55 14 % 2 7 8,5 1,18 16,73 4 0,18 0,054 0,00 8,50 1,36 19,28 13% 3 8 8,5 1,22 17,29 1 0,06 0,072 0,00 8,50 1,28 18,14 13% 4 9 8,5 1,35 19,14 2 0,1 0,060 0,00 8,50 1,45 20,55 7% 5 6 8,5 1,13 16,02 5 0,22 0,053 0,00 8,50 1,35 19,14 35% 6 7 8,5 1,18 16,73 3 0,13 0,052 0,00 8,50 1,31 18,57 36% 7 8 8,5 1,23 17,43 2 0,09 0,054 1,10 9,60 1,32 16,57 37% 8 9 8,5 1,35 19,14 0 0,00 0,000 1,10 9,60 1,35 16,94 48% Fig. 5 Test Results test n D.O.D Discharging Charging? Ah? kwh Partial Efficiencies % Ah kwh Ah kwh Ah n - Ah n-1 charg./ discharg. kwh n - kwh n-1 charg./ discharg.? Ah? Ah? kwh? kwh / / / / / / / / / / / / / / / / / /

12 / / Fig. 6. Charging and discharging partial efficiencies. 12