Zinc Air Battery Stack

Zinc Air Battery Stack

 

The validated unit cell proved the operating principles of the battery over extended time periods and facilitated gradual scale up in electrode. However a flow battery requires a battery stack where up to 200 unit cells are stacked in series with positive and negative electrodes either side of a common bipolar electrode. The fluid systems are normally in parallel and this was the case with the Powair battery even though one of the fluids was air. The unit cells in series are clamped between rigid metal plates to seal the assembly using gaskets or o’rings. Balance of plant items are required to allow the stack to function as a battery (pumps, tank to store the liquid electrolyte, compressor to provide air to the air electrode, sensors, control system etc.).

The initial planning of the project envisaged an air electrode similar to that in a PEM fuel cell i.e. a thin membrane with catalyst on the air side facing a gas diffusion layer / current collector / gas delivery channels with a liquid flow channel on the other side. This style of electrode proved very difficult to develop and the design settled on for the air electrode comprising a catalyst / PTFE mixture which was compressed into nickel foam. This was 0.75 mm thick and fairly robust but was more porous than a thin polymer membrane and cannot be easily bonded to window frame gasket or sealed by o’rings making sealing more challenging.

The general design requirements were:

  • A4 electrode size
  • Bipolar stack of unit cells which consist of a liquid flow channel into which a zinc electrolyte is pumped and zinc is electrodeposited (on charge) or stripped (discharging) onto an electrode plate. On the other side of the conducting plate (bipolar electrode) is the air electrode and the plate must supply gas (air or oxygen) to the back of the air electrode to enable oxidation or reduction of air to occur inside the air electrode plus it must conduct electricity into the air electrode.
  • The air electrode must be pressed onto the current collecting bipolar plate to make good electrical contact and some form of structure in the liquid flow channel to transfer this load was designed.
  • Minimisation of leakage currents along the flow distribution channels.
  • Current collectors at each end of the stack to allow electrical connections.
  • Clamp plates at the end to supply uniform pressure to the stack to allow sealing.
  • Materials of construction which are compatible with the high alkali (8M KOH) and high temperature (60’C) – PP was chosen for the frames and EPDM for the seals.
  • There are a number of requirements for sealing:
    • Bipolar plate into the plastic frame – o’rings, potting using an adhesive or gasket seal.
    • Air electrode
    • Face between the frames.
    • Around the manifolds
  • Easy to modify and allow repeated assembly and disassembly during the development process.

 

The layout of the system is shown in the process and instrumentation diagram for the system, Figure 9.

Figure 9  Process and Instrumentation Diagram for the Stack Tests

Figures 10a 10 Frame stack on a bund with a glass catch pot on the air return from the cell.

Figures 10b  Close up of the top of the cell showing the tie bars and springs that were used to maintain uniform compression.

 

Battery testing was carried out with the assistance of DNV GL / KEMA. An initial observation was that the zinc air battery also exhibits a short term Zn-Ni response, see Figure 11 due to redox cycling of the nickel in the catalyst and nickel foam of the air electrode which is more energy efficient (75 to 80% energy efficient) than the longer term zinc air response (40 to 60% efficient). It is very limited in capacity but could be used to give good efficiency for short term response applications.

The changes in the zinc air response of the battery to changing operating conditions and between runs can be easiest summarised in Butler-Volmer plots of cell voltage against log current (Figure 12) during charge and discharge, the difference between the upper and lower curves represents the losses.  The response is dominated by the air electrode as there is little difference between the deposition and stripping potentials for zinc. The charge curves (the upper curves) change very little between the different runs and between different air electrodes. This is not surprising as it the air electrode reaction is oxygen evolution from water splitting which will not be mass transport limited. However considerable differences are observed between runs for the discharge (oxygen reduction reaction) which is known to be limited by mass transport of oxygen into the air electrode to the three phase reaction zone. It is noted that the electrode performance deteriorates during cycling and this is much more rapid above a current density of 30 mA/cm2, with the discharge voltage dropping and a lower maximum current. There was also differences between fresh electrodes. These differences are not thought to be due to the air electrodes, as testing of small sections cut from different large electrodes appeared relatively consistent. It is thought that the differences were due to the variations in the contact resistance achieved between the air electrode and the current collector.

Figure 11  A short (700s) charge and discharge of a two frame stack Zn-air battery at 12 A showing the initial zinc nickel response.

Figure 12 Butler Volmer plots of a 2 frame battery stack operated with KEMA showing the effect of switching from oxygen to air

 

Figure 12 shows the difference between air (20% oxygen) and oxygen on consecutive runs on the same air electrode. The curves are identical on charge but the air curve drops off more rapidly and beyond discharge is not possible at a much lower current (representing about 40 mA/cm2) than with oxygen which can be discharged up to almost 150 mA/cm2 in some cases.

Almost identical behaviour is observed for the 5 and 10 frame stacks. Example charge discharge cycles are shown in Figures 13-14 for 5 and 10 frame battery stacks.

Figure 13  Transition of the cycling of a 5 frame stack from 12 A (20 mA/cm2) to 30 A (50 mA/cm2).

 

Figure 14  Cycling of a 10 frame stack at 30 A (50 mA/cm2) and 60 A (100 mA/cm2).

 

Battery stack conclusions and achievements

  • A bipolar electrode stack for A4 sized Zn-air has been designed and commissioned along with balance of plant with up to 10 frames in place. This is the largest metal air battery stack of which we have knowledge.
  • The stack operated at up to 120 A (200 mA/cm2) during charge and 90A (150 mA/cm2) in discharge.
  • Peak power obtainable during discharge was in excess of 0.5 kW short term and 0.33 kW long term.
  • The stack performance deteriorated during cycling due to changes in the air electrodes. This was faster with the larger stacks which may be a consequence of the series current connection whereby weakness in one electrode limits all the electrodes.
  • Using air instead of oxygen the output power reduced below that achieved with oxygen after a current density of 20 mA/cm2 and by the limit value of 50 mA/cm2 for air, it was almost 20% lower. Conversely the power using oxygen was only limiting at 100 mA/cm2, indicating the mass transport limitations of the current air electrode structure.
  • While the coulombic efficiency was quite high (>80%), the voltage efficiency was around 60% at low current densities and 40-50% at high current densities. Thus the round trip energy efficiency was quite low and potentially limits the commercial application of the system.

 

 

Consortium Websites

CEST logo

E-on logo Fuma-Tech logo
GreenPower logo DNV KEMA logo
University of Seville logo University of Southampton logo

 

Funded by

FP7 Logo