Energy storage #3

So it has been some time and the cells have been sitting in their crates in the garage. The main reason for this has been the long delay in getting the chassi parts from the water-cutting and machining, even the anodising took a long time. Then, workload has prohibited time in the workshop as desired. No damage, I kept regular watch over the cell voltages and they changed unmeasurably during the time, in fact they all stayed within +-10mV from each other.

This view is from bottom and the 12 mm AL plate that acts as support for the cells.  Some machining was required partly from poor design om my part, partly because the anodising builds thickness. As for the design issue, the cells need to be held tightly in a structure that can take increased pressure, but not so tight the cells are squashed. It turned out that a washer between the square rods created just exactly what was required. The reason for my poor design was that I designed the cages befor the batteries had arrived and I had no opportunity to verify the tolerance of the cells, as it turned out they where slightly thicker than the specs indicated. A better design would have been to make the bottom plate fit just like the square connecting rods. Next time...

Here can be seen how the cell support bar rests nicely in the groove in the side panel. The problem arose when I hade to add the shim washer to the connecting rods, thus offsetting the threaded holes. Widening the holes cured the problem. 

Here is the first assembly test of the first of the two batterpacks. Each pack is a  4P4S configration thus giving 12 volts nominal at 400Ah. These are then connecting in series to form the 24volta as required. The basic design of the battery cage is simple, two 12mm AL endplates are held together by 2x4 square AL rods. The dimension is 15x15 mm and have an M8 thread in each end. The reason for making two packs is that they are made to fit the Artemis, also each pack  weighs approximately 60kgs, the limit of how much two healthy guys can handle. 

Here is detail of sideplate and the stainless washer acting as shim. To the left is the M8  round head Allen bolt. I only use AL and stainless, never mildsteel on the boat.

Closeup of two cells negative terminals. These where made of copper with AL nut and hade some sort of  layer on them. Careful sanding brought out fresh copper that was then cleaned with alcohol. 

The only dirty job was abrading the terminals. I used a  scotchbrite type abrading pad, this was a lot more effective than  sandpaper. Both negative and positive terminals where abraded, the positive are aluminium. All connection surfaces where abraded not forgetting the connecting strips. All was done in one process so air did not have time to produce new oxide layer. 

Winston specifies no more than 20Nm of torque on the M8 teminal bolts. Therefore  I used finger power to ensure the  threads where not sticking, then tightening with torque wrench. You can also see the copper plate (covered by yellow protective tape) that connects the two groups of four batteries. It is very important to ensure that no strain is put on the terminals. Hence, inter-terminal connection is carried out after the packs are finally assembled in their cages.

Two packs now mechanically assembled but with connecting strips yet to be put in place. Having a mindset of respecting the incredible amount of energy available even thought the cells are not fully charged is essential in order to avoid accidental shorts. I was diligent in covering the areas not being worked on and increased the cover as connection progressed. 

When the BMS cell PCB arrive they dont have the negative terminal wire connected. The reason for this is that this modular system can be mounted on a great deal of different cells with varying terminal distances and connecting threads. The manufacturer recommends using a single solid wire, whilst this may look good as you can route it neatly, it has no place on a boat. All single connecting wire I use are of the Huber-Suhner brand. They have double insulation that is tough and can stand heat well and are multistranded plus, of course, tinned. The picture shows both crimping and soldering. Typically I avoid using these insulated type of lugs as the crimping is not satisfactory. Professional crimping produces a gas-tight seal and this can only be achieved using un-insulated lugs.

These are the BMS PCB´s. Made by Dutch company 123Electric.  It is of the
distributed kind where each group of cells has one of these boards connected across the terminals. They are thus selfpowered by the cell. All PCB´s are interconnected by a single wire data bus. At the first and last position are located in and out PCB´s that perform the communication and external device control. The system is quite clever and the distributed architecture makes the system scalable to large series coupled battery packs. There are many choices of BMS out there, most are centralized architecture meaning that there is a wire from each battery terminal to a central unit. There are pros and con´s with eavh of these.
The holes for PCB BMS had to be drilled to accomodate the M8 bolts. There are other ways to do this by using center drilled holes in the bolts.

The BMS allows for separation of charge and discharge current paths. This makes sense in an automotive application but becomes somewhat useless in a service battery system where charging is done by many sources and discharging by even more. The picture shows two 200A current probes and two heavy duty 200A contactors (relays). I will be installing this as if it was a car application to begin with because I will be charging via one single charger. Later, the contactors will be replaced by smart diodes thus allowing to separate in and outgoing current to the packs. The benefit from user sense with paths separated is that the BMS123 phone application (Bluetooth) will clearly present what goes in and what goes out of the battery. 

The two contactors mounted on one of the connecting rods. Also seen is the current probe and 300 amp fuse. As for cable dimension I used no more the 35mm cable. This may seem a bit odd as some cables on Artemis are 70mm. But, keep in mind that cable resistance is a function of cross-section area and length. The pack cables as short. 

Even before final install of contactors and current probe I wanted  to charge and balance the cells. Initially I installed a manual disconnect between the packs, this is preferred method as it means that both the negative and positive side of the packs become floating. But this was installed on one of the cables connecting the packs. Thus introduced a dissymmetry as the electrical center of pack #2 is in fact at the other end of the pack. The disconnect switch was later moved to this point. With the added benefit that the current pickup point between the interconnection is made so that eventual voltage drop in the 4P sub-packs is minimized. Working with such high currents it is essential to always consider that very small resistances play a role, the also add up. Every milliohm matters!

This is a 24volt 20A charger sourced from EV-power.eu. A Chinese job at decent price and it works well.  It is a charger settable for LFP chemistry. It can be remotely controlled by a BMS, but it turned out that when the BMS123 turned off the charge contactor, the charger went into idle. Very nice. The simpler, the better.

The Anderson power connector system is very clever and even more simpler. I installed  this plug in the charger wire.  

In order to charge and then discharge this quite large battery bank to both test and also activate the cells we need a load. Lacking the professional gear that essentially are bi-directional powersupplies that get power from the grid and then put it back when cycling the pack, I am forced to dump the energy as heat. An easy way to do this is using a 230Volt heater. Hence this cheap 3000watt inverter to produce the voltage required. The rate of discharge is 0,1C meaning for a 10kW pack that 1kW of power needs to be dumped for around 10 hours.  

This is the crate that has been used to ship "stuff" to and from the boat in southern Italy. Will soon be used to ship the two batterypacks to their new home.

In the next blog on energy storage I will talk more on the BMS, why we need supercaps, some add-ons to the pack and other insights gained building this pack.