Concept for electrically-driven work machines Jussi Suomela, Jussi Puranen, Juha Pyrhönen

Abstract Wheeled and tracked work vehicles (work machines) seem to be very comparable to transportation vehicles when it comes to their electrification. However, the variety of work machines of different sizes and for different purposes is huge, and, because of this, the motivation and requirements for their electrification differ greatly from those for road transport. In this study selected products from four Finnish work machine manufacturers were analysed in order to find out the common denominators in their motivations and electrification requirements. According to the requirements, a general design concept for the subsystems of an electrically-driven work machine was created. Despite their very different intended uses, from a harvester to a straddle carrier, certain common denominators were found. The concept includes the defined parameters/types of the electric subsystems, such as the power bus, buffers, driving motors etc. The concept did not commit itself as regards the energy source but the “basic principle” of the study was that in the future the problem of on-board electricity supply will be solved with a fuel cell. Keywords: EV, off-road, electric drive.

1

Introduction

Fuel cells and hybrid power systems are a hot topic in the car and bus industry. The main motivations are to have zero-emission vehicles and reduce fuel consumption and thus the dependence on oil. Electric drives will also generate other “side benefits” such as less noise, better controllability, a longer lifetime etc. Wheeled work machines are very comparable to transportation vehicles when it comes to their electrification. The motivation aspects with work machines and transportation vehicles are to some extent the same, as is the main problem – how to carry enough electrical energy on board. However, the variety of work machines of different sizes and for different purposes is huge, and because of this the motivation and requirements for electrification differ greatly from those for road transport. Among existing work machines the size really matters. The biggest machines are more often electrically powered. Rail-based machines, such as different types of cranes, are naturally electric because of the ease of supplying power, but freely-navigating machines such as gigantic dump trucks, straddle carriers, and LHDs (Load Haul Dump machines) (Fig. 1) are also often electrically powered. The three last-mentioned machines are a good example of how different the machines and the motivation for electrification can be. LHDs are cable-supplied and the main motivation is to have zero emissions, which reduces the need for ventilation underground in mines. Dump trucks and straddle carriers are both diesel-electric. In a dump truck the hybrid system saves fuel and reduces emissions as a result of the constant-speed diesel. Electric drives provide better controllability, the possibility of using a trolley supply, and electric braking with regeneration. Additionally, the maintenance costs are smaller than with a traditional machine with mechanical transmission. Because of the complicated structure of the straddle carrier, the electric transmission is easier to build, but the main motivation is the reduction in maintenance costs. According to Kalmar, the extra components of the diesel-electric straddle carrier make it more expensive than a diesel-

mechanical one, but in a couple of years this is paid back in lower fuel consumption and longer intervals between maintenance periods.

Figure 1: Electric work machines: Gigantic dump truck (Liebherr, adopted from [1]), Straddle carrier (Kalmar), and LHD (Sandvik Tamrock) At the other end of the size scale small indoor forklift trucks are mainly electric, but the wide area between these two poles consists mainly of diesel-driven (mechanical/hydraulic) machines. In the study the task was to find out the motivation, common interests, and solutions for electrification among four work machine manufacturers with very different products. The cooperation and common interest are very welcome because of the small volumes involved. The participating manufacturers may be the biggest ones in their own market but their volumes are tiny when compared to those of transportation vehicles, especially passenger cars. The participating manufacturers were Kalmar Industries (Kone Cargotec), Plustech (Timberjack/John Deere), Patria Vehicles, and Sandvik Tamrock. Examples of their machines are presented in Fig. 2. In practice the idea was to define a concept for the electrical subsystem of a future work machine based on fuel cell energy. The study concentrates on the electrical subsystems, supposing that enough electric energy is available on board. Thus, in addition to fuel cells, any feasible energy supply, such as a gen-set or a trolley, can also be used. The study was targeted to the following questions: • Motivation: Why? Benefits of electrical drives, lifetime, loading, cost of operation, manufacturing costs, environmental aspects, etc. • Special characteristics of work machines: Typical work cycles (power, torque, speed as function of time) and environmental requirements. • Benchmarking, commercially-available electrically-powered vehicles and their parameters: work machines, cars, buses, trams (ships, trains) • Electrical subsystems for electric vehicles: how subsystems can fulfil the special characteristics of work machines • Evaluations and visions: o What the bottlenecks in work machine electrification are and how they can be solved o Possible future standards: energy bus, data connections o Proposal for concept of electrically-driven work machine

Figure 2: Products of participating manufacturers: LHD and dumper by Sandvik Tamrock, reach stacker and straddle carrier by Kalmar Industries, AMV by Patria Vehicles, harvester and forwarder by Timberjack

2

Demands of electrification

The topic was to find out the special needs for the electrification of the machines of each participating work machine manufacturer. The main concerns were: motivation, performance demands, drive cycles, environmental demands, and existing electric work machines.

2.1

Motivation

The motivation was studied with a questionnaire among the participating manufacturers. As expected, the level of readiness for fuel cell technology was one of the main motivations for electrification. Other wellranked matters were emissions, the possibility of using wheel motors (structural freedom), and controllability. The opinions of the participating manufacturers are listed in Table 1.

Motivation Readiness for fuel cells Structural Emissions Noise Controllability Need for maintenance Heat emissions Efficiency Other Other

Sandvik Tamrock Important Wheel motors Important Yes

Yes

Patria Yes Wheel motors Yes Important Yes Yes Important

Plustech Yes Yes Yes Yes Yes

Kalmar

Yes Yes Yes Important Yes

Synergy Power production

Table 1: Motivation of manufacturers (“Yes” means basic interest)

2.2

Performance demands

The vehicles studied were of very different types. Therefore the idea was to find out common denominators which are critical for electrification. At the beginning the aim was to study all types of actuators in the target vehicles. However, it was soon noticed that in high-force linear actuators, such as boom control, and in other very high power/weight (and size) ratio applications, such as e.g. harvester saws, hydraulics is still the best and the only solution. It was also considered that replacing a diesel engine with an electric motor in a hydraulic pump or a low-power hydraulic actuator with an electric actuator is relatively trivial. The manufacturers had already had experience of these kinds of replacements in their existing machines. Because of these facts the study was targeted to driving actuators, which in most of the cases consume most of the power of the vehicle. The objective was to find out if it is possible to use electric motors for driving with a constant transmission rate (wheel/hub motors) without unfeasible overdimensioning of the motor power. It was noticed that work machines have an extremely high torque ratio between low- and high-speed driving. In normal transportation (passenger cars, etc.) the torque ratio is around 5-10, and in heavy transportation (trucks etc.) around 20. Among all the vehicles studied only Kalmar’s straddle carriers fit into the transportation class. Others need torque ratios up to 30. The transportation torque ratios (from 510 up to 20) are relatively easy to reach with electric motors but the ratio of 30 is a difficult one. Drive motor selection is studied in Chapter 3. Additionally the studied machines are typically running long periods with (near) full power, which limits the use of peak buffers.

2.3

Environmental demands

The electrical devices needed, especially inverters, are traditionally made for industrial use and therefore their environmental specifications usually cover only a narrow area, while all the work machines studied have very tough demands. On the other hand, transportation vehicles have more or less similar requirements. The environmental requirements are listed in Table 2. It can be seen that standard industrial devices do not fulfil these requirements. However, there are no physical limits to the manufacture of inverters, motors, and other electrical devices for the conditions required. It was noticed in the benchmarking that some equipment manufacturers already offer these kinds of heavy-duty, liquid cooled inverters and motors for mobile machines. To fulfil these requirements, in practice, both the inverters and driving motors have to be sealed and liquid-cooled in order to keep dust and water away and to ensure sufficient cooling under all conditions. The vibration durability also has to be considered in the design of the equipment. However, all electrical subsystems can be, and most of them have been, used under these conditions without problems. The most critical subsystem will be the fuel cell, because of the environmental temperatures (freezing at low temperatures and cooling at high ones) and impurities in the incoming air. Batteries also have problems at low temperatures. Environment Operating temperature Atmospheric pressure Humidity

Sandvik Tamrock -20…+55

Patria -46…+49

Plustech -35…+45

Kalmar -40…+40

-3000 m…+4500 m High

70…106 kPa

Conventional

Conventional

High

0…80%

Vibration

1.5 m/s^2, 2-10 g

10-100%, driving wheels under water Present

Continuous

Corrosive agents Inflammable materials

Sulphur, salt water

Explosion hazard IP-class Foreign objects

Requirement for extinguishing system In coal mines 67+ pressure cleaning-tolerant Dust

People Serviceability

Present Must be good

Modularity

Favourable

Table 2: Environmental requirements

Ice-control salt, sulphur dioxide Dry terrain

Oil, wood dust

0.2…0.3 g in frame Industrial climate Oil/fuel

EX-protected

Not actual

No

68

Rainwatertolerant

IP 54

Rocks, stubs, barbed wire, dust Present Small maintenance in vehicle, large maintenance out of vehicle Separable rims

Dust

Possible in forests

Desirable

Not close Maintenance interval 1000 h, easy maintenance Good

3

Traction motors

A common denominator for the machines studied was the need for a high torque ratio. To clarify whether it is possible to achieve this ratio without remarkable over-dimensioning of the motors or shiftable gears, which would be a problem in the case of hub- or wheel-specific motors, traction motors were studied more carefully.

3.1

Suitability of different motor types for electric vehicles

In the past, the DC motor, with its excellent control characteristics, was the only controllable motor type, but nowadays it has, to a large degree, been replaced by different AC motors. AC motors are maintenancefree, more robust, and, with modern inverters, it is possible to achieve dynamic performance comparable to that achieved with DC drives. There are basically four motor types that are suitable for use as traction motors in electric vehicles: a separately-excited synchronous motor, a synchronous reluctance motor, an induction motor, and a permanent magnet synchronous motor. Excellent dynamics and a long field weakening range would be possible with a separately-excited synchronous motor, but since there exists a requirement for a separate magnetization device, it is far too expensive a solution to be used here. Additionally, the mechanical construction of the motor does not allow operation at speeds beyond several thousand rpm. Synchronous reluctance motors have an extremely simple rotor construction and they are used in some special applications, but even then they must be custom-made, because at the moment they are not available as a standard product. Thereby there are basically two motor types left; the permanent magnet synchronous motor (PMSM) and the induction motor (IM). The control theory for both of them is very well-known, they are both widely available from numerous manufacturers as standard products with powers ranging from hundreds of watts up to the MW range, and their prices are quite moderate compared to other motor types. In electrically-powered vehicular applications designed for transportation, such as passenger cars, buses etc, traction torque is usually produced by a permanent magnet synchronous motor (PMSM). The main advantages of a PMSM in vehicular applications are: • high power density • high efficiency • accurate speed control without speed sensor possible These characteristics make the PMSM especially suitable for lighter vehicles, where the torque requirements of the traction motor are relatively small. The sizing of an electric traction motor must be performed with respect to two factors; first, the motor must be capable of producing adequate torque while starting and for acceleration, and second, the operating range of the motor must extend to the maximum speed defined by the application. With electric vehicles designed for transportation, the starting torque is usually the maximum torque the motor must produce. The torque ratio between low- (or zero-) and maximum-speed operation (denoted as R) defines the sizing of the motor. For transportation vehicles, such as passenger cars and buses, this ratio varies between 5-10. This means that the nominal point of the motor can be located in the high-speed region, and with a moderate gear ratio, enough torque is available for starting and acceleration. A heavy working vehicle is a totally different kind of application, because extremely high torque is required at zero speed and at low speeds. For example, with the load haul truck the loading conditions as follows: the bucket is pushed under the pile of rocks with the assistance of the traction wheels, and speed during this process typically varies between 0-2 rpm. The torque required during this operation is typically 15-30 times the torque required while driving at maximum speed. This makes the sizing of the motor for heavy working vehicles very difficult; the motor must be capable of providing extremely high torque at

very low speeds, and also relatively high torque while driving at high speeds. A typical torque-speed curve of an LHD is illustrated in Fig. 3. In this specific case the torque ratio R is approximately 16, but, in the worst case with LHD and with the AMV, the corresponding ratio is over 30. LHD torque-speed curve 80000 70000

Torque [Nm]

60000 50000 40000 30000 20000 10000 0 0

10

20

30

40

50

60

70

80

Rotation speed [rpm]

Figure 3: Torque-speed curve of the LHD. The torque ratio between low-and high-speed operations is approximately 16. With the PMSM, the nominal point of the motor would have to be located near the maximum speed, because exceeding the nominal speed of the PMSM is very difficult to carry out, because the field weakening characteristics are very poor. The flux produced by the stator currents (armature reaction) has a minor effect on the total flux of the machine, which is, to a large degree, determined by the constant flux produced by the magnets, which means that if field weakening is required, a high demagnetizing current component is required on the stator. Further on, when most of the total stator current is used to demagnetize the machine, there is only a little current left for the machine’s q-axis to produce torque. The point at which all the current available is on the machine’s negative d-axis to produce the demagnetizing flux and there is zero current left to produce torque is reached soon after the nominal speed. This is the maximum speed of the motor, where the torque, and consequently the power, drops to zero, and with PMSMs with a small armature reaction, this speed is usually 1.1-1.2 times the nominal speed. The nominal speed of the PMSM is located near the maximum speed, and the motor produces less torque at low speeds, and gearing with a relatively high gear ratio is required to step the torque adequately for lowspeed operation. Also the overloading capability of commercial PMSMs is moderate, the pull-out torques are usually below 1.5-2 times the nominal torque.

3.2

Induction motors in heavy traction drives with high-speed operation

With modern frequency controlled induction motor drive, where the torque and flux are controlled separately, field weakening is a common technology, and speeds up to 4-5 times the nominal speed are possible. This means that the induction traction motor can be sized so that its nominal point is located at a lower speed, and the field-weakening region is also utilized. The optimised torque capability guarantees that the motor will produce adquate torque in the field-weakening region, too. Fig. 4 illustrates the operating curves for both an IM and PMSM with approximately equal power. It can be seen that with an induction motor with equal power, much higher torque is available during low-speed

operation than with the PMSM, because the nominal point of the PMSM must be located at a higher speed in order to cover the speed range, because the field weakening range of the PMSM is extremely short.

Figure 4: Comparison of operating curves between induction motors and permanent magnet synchronous motors. The field weakening range of IM is remarkably higher, which means that the motor’s nominal point can be located at a lower speed and more torque is available during low-speed operation. The mechanical construction of the rotor of an induction motor is very robust, and rotation speeds up to 10,000 rpm are possible. In order to produce the required torque during field weakening operation, the pull-out torque (i.e. the maximum torque) of an induction motor must be high enough, because the torque of an induction motor decreases inversely proportionally to its speed during field weakening, and, if the load torque exceeds the motor pull-out torque, breakdown occurs and the drive must be stopped. The benefit of an induction motor is the fact that with optimal design, its pull-out torque can be remarkably increased. This subject was studied in [2] and it was shown that the pull-out torque can be relatively easily increased up to over five times the nominal torque. The disadvantages of such a design were increased losses and increased torque ripple. With liquid cooling, however, the losses can be dissipated. Usually the coolant medium is water, but gearing oil could also be used. Torque ripple is probably not a problem either, because the load inertias are extremely high and high-frequency torque ripple is effectively filtered out. Comparing to PMSMs induction motors can also be operated with higher temperature because of the demagnetisation effect caused by the temperature in the PMSM. This is important when operating in high environmental temperatures.

4

Energy/power system

The main task of the energy system is to provide enough good-quality electrical energy to the machine actuators under all conditions. The system structure should be suitable for any kind of supply (gen-set, fuel cell, trolley, battery, etc.) and their combinations. It should also support the use of different kinds of buffers, such as batteries and super-capacitors. Typically the main loads in a work machine are the driving and hydraulic pump motor(s). The basic structure of an electrical energy system is illustrated in Fig. 1. The main parameters of the energy system are the current type (DC or AC), voltage level, and buffering, if needed.

4.1

Voltage level

In mobile work machines the voltage should, normally, not exceed 1000 V, which is the limit for highvoltage systems. On the other hand, the machines studied need high power levels up to 400 kW, i.e. the voltage should be as high as possible so as to avoid high currents. It seems that the practical upper AC limit is 690 V, which is the maximum for most of the standard equipment (inverters, motors) and the lower limit depends on the equipment used but does not go below 300V. In practice, the best voltages are 400 or 690 VAC, which are the nominal voltages of standard equipment. In the case of a DC bus the inverters are connected from the intermediate circuit and the bus voltage must be at least 1.35 higher than the AC voltage desired.

4.2

Voltage variation

Commercial inverters typically allow input voltage variation between +/-10%. The minimum DC voltage level is 1.35 x nominal AC output and the maximum (depending on components) is about 700 VDC in 400-VAC inverters and up to 950 VDC in 690 VAC inverters. Even though the variation in the intermediate circuit can be very wide, direct fuel cell supply will be a problem. The load response of a PEM fuel cell is illustrated in Fig. 5. It can be seen that when the load increases from zero to the nominal point the output voltage decreases by 50%. This is too much, even in the case where a 690 VAC inverter is driving a 400 VAC motor.

Figure 5: Load response of a PEM FC

4.3

Bus-type AC/DC

The existing power supplies in high-power machines are mainly of the AC type (network or generator). All types of fuel cells provide DC and all the possible buffers are also of the DC type, except flywheels. Technically, both AC and DC supply buses are feasible. However, considering the cost and functionality of the whole system, the DC solution seems more practical, especially in the case of a DC supply and when buffers and/or braking energy exploitation are applied. The differences between the bus types are illustrated in Fig. 6.

FC DC

AC

DC

DC

AC

AC

Umbilical or trolley

Umbilical or trolley

G

AC - bus

FC

B

M

Braking resistor

Braking resistor

M

AC

AC

AC

M

M

AC DC

DC - bus

AC Braking resistor

AC AC

Braking resistor

AC AC

AC DC

G

DC

DC

DC

DC

AC

AC

AC

AC

M

M

M

M

Braking resistor

B

Figure 6: Electrical systems with AC and DC bus. Note: batteries, capacitors, and fuel cells might need DC/DC converters between the DC bus The main differences between the systems are: • braking energy recovery needs either a DC bus or special (double) inverters • because of this, each driving motor needs its own braking resistor in an AC solution • all feasible buffers have a DC interface

4.4

Buffering

4.4.1

Why buffer?

In a work machine buffering can be needed for the following reasons: 1. To provide short-time peak power Among the machines studied only AMV and LHD would benefit from short-time power peaks, mainly because of power-intensive auxiliary equipment. Even if the driving motors of the machines studied are highly overloaded at low speeds, the power is not overloaded, average power is near the maximum power, and the maximum power is used for relatively long periods (dozens of seconds to minutes) continuously. This is a clear difference when these machines are compared to transport vehicles. 2. To stabilize the bus voltage In this case the buffer mainly filters the electric noise from the bus. Compared to the previous situation, the peaks are very short (only ms) but the response has to be fast in order to keep the voltage level stable. Especially in the case of fuel cell or gen-set, buffering helps to keep the bus voltage stable during load changes. 3. To store possible braking energy If the machine is not connected to the public network, the braking energy cannot be recovered without suitable buffering or simultaneous use of another actuator (such as lifting in the case of a straddle carrier). The amount of braking energy is totally dependent on the task and can be very high in some applications. If buffering and other loads are not enough to consume the braking power, separate braking resistors have to be used. 4.4.2

Suitable buffers

Electricity can be stored in electrical form only in capacitive or inductive storages, which have, because of the physical facts, a small energy/size ratio. If more capacity is needed the form of the energy has to be changed. The most typical forms are chemical energy (batteries), mechanical energy (flywheels), and

potential energy (water storage). Taking into account the mobile machines and existing technology on the market (now and in the near future), the most suitable buffer storages and their main properties are: 1. Different types of batteries (Lead-acid, NiCd, NiMH, LiIon, LiP) • High energy density • High efficiency • Stable output voltage • Limited lifetime 2. Super (or ultra) capacitors • Long lifetime • Simple structure • High efficiency • Very high power density • Poor energy density • Voltage drops 3. Flywheels • Long lifetime • Poor power density • Poor energy density Because of the large size of existing flywheels, the only feasible buffers in a work machine are supercapacitors and batteries. Because of their high power density capacitors provide a fast response to very short peaks. Batteries can buffer longer power peaks and store braking energy as a result of their relatively high energy density.

4.5

System control

An electric work machine will be totally under computer control, as many traditional machines already are. The possible wheel-specific motors also bring a greater degree of freedom in the matter of control. For example, an eight-wheel AMV would have steering + 8 separate wheel speeds to control instead traditional steering + speed. This provides better performance but also complicates the control. The distributed electrical loads will make both energy/power and temperature control more complicated.

5

Conclusion and future work

It is not possible to define an unambiguous concept for an electric work machine. However, to conclude this study briefly, which is a conclusion itself, a very raw definition is made. One must remember that the study is based on a limited amount of case examples. Electric work machine concept: • Power supply: fuel cell, gen-set, trolley or combination • Main power bus: DC • Bus nominal voltage: 550 VDC or 950 VDC • Driving motors: asynchronous, water-cooled, 400 or 690 VAC • Inverters: vector control, water-cooled, 400 or 690 VAC • Buffering: supercapacitor (bus stabilization) and battery (energy storage, if needed)

The study is being continued at the moment with a fuel cell-related project, “Concept for fuel-cell driven work machines”, which includes the design of a fuel cell-powered electric work machine and an experimental prototype based on an existing work machine and a commercial PEM fuel cell.

References [1] Liebherr Group, Mining Trucks, available online at http://www.liebherr.com/me/en/47597.asp [2] J. Puranen and J. Pyrhönen, “Analysis of a pull-out optimized induction motor in heavy traction applications”, International Conference on Electrical Machines, ICEM 2004, Cracow, Poland, Sept. 2004

Authors Jussi Suomela, Dr. Tech., Senior Research Scientist, Intelligent Machines and Special Robotics Institute, Helsinki University of Technology, P.O. Box 5500, 02015 HUT, Finland phone: +358 9 4513301, fax +358 9 4513308, e-mail: [email protected] Jussi Suomela received his M.Sc. degree in electrical engineering from Helsinki University of Technology (HUT) in 1992. Since then he has carried out several research projects in the area of mobile robots and intelligent machines. From 1995 he has worked as project manager. He received his doctoral degree from HUT in 2004.

Jussi Puranen, M.Sc.(tech.), research engineer, Lappeenranta University of Technology, P.O. Box 20, 53851 Lappeenranta, Finland. Phone: +358-5-6216766, Fax: +358-56216799, email: [email protected] Jussi Puranen received his M.Sc. degree in electrical engineering from Lappeenranta University of Technology in 2002 and has been working towards his D.Sc. degree since then. His research interests are electrical machines, especially AC machines in traction and high-performance drives.

Prof. Juha Pyrhönen, Lappeenranta University of Technology, Department of Electrical Engineering, P.O. Box 20 FIN-53851 Lappeenranta/Finland. Tel: +358 5 621 6706. Fax: +358 5 621 6799. E-mail: [email protected]. Juha Pyrhönen (1957) received his M.Sc. degree in electrical engineering from Lappeenranta University of Technology (LUT) in 1982. He received his Licentiate of Science (Technology) degree from LUT in 1989 and his degree of Doctor of Science (Technology) from LUT in 1991. Juha Pyrhönen became Associate Professor in Electric Engineering at LUT in 1993 and Professor in Electrical Machines and Drives in 1997. He is currently working as Head of the Department of Electrical Engineering. He is engaged in electric motor and electric drive research and development. He is also the leader of the Carelian Drives and Motor Centre, which, in cooperation with the Finnish ABB Company, is developing new electric motors and drives. Synchronous motors and drives, switched reluctance motors and drives, induction motors and drives, solid-rotor high-speed induction machines and drives, as well as active network bridge control, are included in his current interests. He is leading the research work of several postgraduate research groups working in the target areas mentioned above.