The Fool on the Hill: Modular electric farm traction

Contents

1. Battery weight, and modularity
1. Geometry
2. Swapping Modules
2. Motors
3. Arrangement
1. Alternate uses for battery module bays
2. Geometry: Front loaders
4. Charging
5. Conclusion

Currently, farm tractors almost all burn diesel, and (with very few exceptions) burn fossil fuel. To reduce the carbon cost of farming, either electric or hydrogen powered tractors are necessary. This document considers electric power.

Battery weight, and modularity

A family of modules is needed to meet a range of traction needs. The smallest useful farm tractor for current farming is about 30 Kw; tractors up to about 200 Kw are in common use. Current lithium ion batteries deliver about 0.25 Kwh per kilogram. To operate at near full power for an eight hour service day consequently requires

However, at peak times of year, particularly harvest, tractors often see a duty cycle considerably exceeding eight hours a day; a full day's duty cycle might well be sixteen hours.

These values imply that for a sixteen hour service day the battery alone would weigh up to twice the total weight of an equivalent fossil powered tractor; which is implausible, partly because soil compression is an issue, and partly because a heavier tractor will use more of its power moving itself. So the batteries need to be modular and to be quickly and easily swappable, including in the field. But it's also implausible that manually swappable batteries could be used, since even a thirty kw tractor with an eight hour service battery has nearly a ton of battery, meaning forty 25Kg modules. I think four 250 Kg modules, each delivering 62 Kwh is more practical, but means that they must be installed either with a fork lift or front loader, or by means of static lifting gear in a shed.

In 2019, lithium ion batteries cost US$156 (about £123) per Kwh, and that price is still falling. So the cost of batteries per 250Kg module is about £7,350. This is also challenging. If you're going to be able to field swap batteries to run sixteen hours a day at harvest time, you're going to need two complete sets of modules, which even on the smallest tractors is eight modules or about £60,000. For comparison, a complete new 35Kw John Deere costs around £25,000. Also, of course, that spare set of batteries is probably only in use for three or four weeks a year, and is otherwise sitting idle; but you can't rely on renting them because most farms will be needing the extra service time at the same times of year. However, against this, the cost of 'fuel' is much lower.

(Note: I'm suspicious of my arithmetic here and will need to recheck it when my brain is working better.)

Geometry

The battery cell size most commonly used in electric cars (including Tesla) is the 21700. It is claimed that this is optimal for energy density and battery life. This is a cell 21 millimetres diameter, 70 mm long, and weighing 65-70 grams each. Cells of this size are available with capacities between 3,000 and 5,000 mAh. It should be possible to package these cells into a module fitting into a 100 mm wide space (i.e. 70 mm cell, 2 x 0.5 mm connection layer, 2 x 14 mm wall thickness = 99 mm). Allowing a 25 Kg weight for the packaging, this gives 3214 cells per module. Allowing 4 mm air-gaps between cells to allow for cooling air circulation, that gives an internal module area of 1400 mm square — i.e. 1.4 metres wide and 1.4 metres high, or slightly under two square metres. In addition to the cells each module must contain

1. The electrical connection to the power bus on the tractor/charger — this is probably one (rather thick!) shielded pin connector for live and a bare metal-to-metal flat connector for ground;
2. The control electronics for the battery;
3. A data connection to the tractor, probably through a near-field connector;
4. A heating/cooling system in order to maintain the cells at an efficient working temperature;
5. An air intake filter for the above;
6. A means by which the module can be suspended from a fork lift, possibly with the aid of some adaptor;
7. Some pretty solid means of locking the module to the spine frame, so that it cannot easily become detached in an accident.

So the overall size of the module is either about 1.5 metres by 1.7 metres by 100 mm, or, by stacking two cells into each column instead of one, about 1 metre by 1.2 metres by 170 mm. The second packaging is what is shown in the illustration above. Either size is bigger than I'd like. I think electric tractors will be more easily adopted if they look something like conventional tractors; the 'one cell per column' module has a cross section area substantially bigger than a conventional tractor engine/transmission, while, if the 'two cell per column' module is used, the overall length of the battery section of a 200 kw tractor is more than four metres, giving an overall length (including operator cab) of about 5.5 metres, without linkages. That seems unwieldy.

With the 'two cell per column' model, however, 50 Kw tractors, using eight battery module positions, would have a 'bonnet' (the module section) under 1.4 metres long which is reasonably conventional and even on the compact side, while a 100 or 150 Kw tractor, using 16 module positions, would have a 'bonnet' 2.8 metres long, which is not excessive.

NOTE THAT the arrangement of cells in the module need not be square; this was a simplification to give me an estimate of the total dimensions required. Also, note that I'm by no means qualified to determine how the cells within a module should be connected, nor what the optimal power bus voltage should be.

Swapping Modules

If battery modules are to be swapped in the field, then a robust field safe electrical interchange between the battery and the power bus is essential. That is a design challenge.

However, it seems to me in the highest degree desirable that the tractor should have the hoist equipment required to swap its own battery modules. It is quite easy to imagine that the module should have a hoist connection (perhaps as simple as a loop) on the top which, when hoisted, first retracts the live pin from the bus and locks it in the raised position, perhaps with an automatic insulating shutter closing under it, then releases latches which secure the module mechanically to the chassis of the tractor, and finally lifts the module clear.

For a tractor to swap its own modules requires under its own power required

1. multiple modules; and
2. that the bus be powered while modules are being swapped.

Obviously this still wouldn't allow modules to be swapped under power if the batteries were totally discharged, but firstly the system should give adequate warning before that situation is reached, and secondly if the modules weigh only 250kg each, then an auxiliary hand winch could be provided for emergencies.

NOTE THAT in the schematic above, I've assumed the battery hoist crane runs on a track based on flanges on the lower side of the spine chassis. This means that with the operator cab in the rear position, it cannot be moved far enough forward to conveniently lift the frontmost battery modules, and with the operator cab in the front position, it cannot be moved back far enough to conveniently lift the rearmost.

Motors

15 Kw hub centre motors for motorcycle use are currently available from China for US$650 each, so it should be possible to source an appropriately robust 25 Kw hub centre motor for less than £1,500 each in quantity. This means each powered pair of wheels — each powered axle — delivers 50 Kw. So a tractor with powered wheels on one axle has an output power of 50 Kw, with two axles 100 Kw, with three axles 150 Kw, and with four axles 200 Kw. There's an issue here in that modern farm tractors don't have suspension, but to keep more than two axles on the ground on uneven surfaces requires some form of suspension. Also, if there are more than two axles, all but one of the axles must have steering (although no universal or constant velocity joints are needed). So it might be better to think, rather than two, three and four axle tractors, instead to think of having 25 Kw and 50 Kw hub motors.

Whether these hubs should include an epicyclic gear system, or just be direct drive, is a matter for someone with more electrical knowledge than me.

It doesn't seem to me that it is necessary the wheels could be swapped in the field, but they need to be able to be swapped on the farm; so there must be an electrical connection which can be disconnected and reconnected without elaborate safety precautions.

The downside of separate motors for each wheel is that there is no differential which can be locked, but the same effect can be achieved through electronic traction control, which is necessary anyway (and easy and inexpensive to implement), so this is not a great consideration. The lack of a clutch, gearbox and transmission shafts, all complex and expensive components on fossil-fuel tractors, is a significant cost saving.

In addition to the hub motors, additional motors are needed for mechanical power take off, and for hydraulics. However, these, too can be modular, and only fitted if wanted (which for most users they will be).

Arrangement

My conception of the overall arrangement, then, is of a box section spine frame, to which at least two axle assemblies are mounted. Battery modules sit across the frame as saddles; the number of battery modules that can be carried is then a function of the length of the frame. It should be possible to run any tractor (for a short time) with just one module in place. The spine frame must thus have a power bus and a data bus; bus connections must safely and automatically be made when a battery is placed onto the frame, and safely and automatically be closed off when it is removed. The axles must also connect to the power and data buses, in order to carry power and commands to the wheels.

The operator cab unit (for non-autonomous vehicles — and although I think autonomous tractors are inevitable, I think that is a whole other research project) must also fit on the spine frame, analogously to the battery modules.

Either end of the spine frame may carry a power take off motor, a three part linkage, or both. The spine frame itself could contain the hydraulic fluid reservoir and pump. The hydraulic pump could be optional, but having hydraulics in the basic package, if only for steering, would be useful.

Probably it makes sense to supply spine frames capable of taking eight, sixteen or twenty four battery modules; obviously, not all module bays need be occupied. For simplicity, I think each of these spine frame units should be identical except for length and number of battery module bays. Thus for a basic 50 Kw tractor you'd probably order a basic 8 or 16 module frame with axles, two 25 Kw wheels, two unpowered wheels, four battery modules, a hydraulic pump, one PTO motor, one three point linkage and an operator cab. You could later upgrade it by adding two more 25 Kw wheels, by adding an extra three point linkage at the other end, by swapping out the 25 Kw wheels for 50 Kw wheels, or by swapping the spine frame for a longer spine frame in order to carry more battery modules.

Alternate uses for battery module bays

Not all uses for a tractor require it to run at full load all day. For lower load uses, it may not be necessary to have all bays occupied by battery modules. Thus it might make sense to have (e.g.) a sprayer tank assembly which fitted into four adjacent battery module bays, taking power from the bus by a standard battery module connection.

If the operator cab module is designed to fit over eight adjacent battery module bays, then by having sixteen actual bays all along the spine frame on a nominally eight battery module tractor, you could have a forward control tractor or a conventional rear control tractor simply by lifting the operator cab module off the frame, moving it to the other end, and lowering it back on. Like the suggested sprayer module, the operator cab module would have exactly the same bus connections as a battery module, and the tractor would be entirely 'drive by wire'.

Finally, an autonomous control module for a fully autonomous tractor need only take up one battery module bay.

Geometry: Front loaders

We have to make provision for front loaders within the modular concept, since they are now widely used vehicles. To make a front loader, a telescopic boom assembly must be situated alongside the main chassis, pivoted near the rear axle and extending forward beyond the front of the vehicle when retracted to its minimum extent. This requires offset axles — the 'standard' axle locates the spine frame in the middle of the vehicle, but the 'front loader' axle must offset it to one side, to allow the boom space on the other side. Also required is an offset cab, which does not extend at all over the boom side. And, of course, it requires the front loader boom itself. All other components would be as described above.

Charging

Most farms, even in Scotland, have enough shed roof area to mount enough solar panels to power all the tractors they use; alternatively, power can be drawn from the grid, although that is currently more expensive per Kwh than agricultural diesel. If solar panels are used, then excess power may be sold to the grid, offsetting the capital cost of the panels. The battery charger unit looks (and is) much like a tractor chassis without the axles: it's a box-section beam with the same battery connections, and battery modules are simply placed on it in order to charge them.

Conclusion

It seems possible with modern technology to produce an extremely modular electric tractor design. The company which produces the best solution to electric farm traction will be very successful. This is an industry Scotland could (just) break into, but we'd have to be fast.

Share on social media