Electrolyzer Integration with Renewable Energy Sources: Part 2 - Challenges.

> Parasitic Losses

Almost all commercial water electrolysis technologies employ the use of bipolar stacks of electrolytic cells. This stacking method enables electrolysis on a significantly higher net surface area with the same amount of current. An N-cell stack with each cell having surface area A has a net effective surface area of N x A. However, the current that needs to be applied for a specific design is only proportional to A. Typical current densities used in AEL are between 0.4 - 0.7 A/sq.cm, while that in PEM is ~2 A/sq.cm.

Therefore, it would seem that decreasing A while increasing N is the best way to reduce current requirement therefore reducing the power electronics cost.

> Reduction in hydrogen production efficiency

However, bipolar cells with common electrolyte channels are prone to parasitic (or shunt) currents.

These shunt currents are undesirable for two reasons:

They cause corrosion of some components of the system, and more importantly, these currents are essentially lost in terms of production of hydrogen. As the ratio of shunt current to input current increases, the efficiency of hydrogen production is reduced.

As the number of cells is increased, the shunt current fraction increases.

So, the number of cells to be stacked in any electrolyzer design is limited so as to keep the shunt current losses minimum at the operating input current. However, this optimization in design is done typically for the nominal current at which the electrolyzer is expected to operate most of the time.

When the actual operational currents are lower (under low load conditions), the shunt current fraction can become significantly higher leading to reduction in hydrogen production efficiencies thereby no longer rendering it cost effective.

> Improper potential balance between the cells

In an ideal stack of electrolyzer cells, where there are no shunt current losses, the entire input current passes through all the electrodes of all the cells, causing equal potentials across the cells and equal hydrogen production.

But in reality, shunt currents always exist due to the conductive electrolyte or due the stack structure typically being made of metal. These shunt currents cause the actual current passing through each of the cells to be lower than what is provided at the end cells. On simulating the equivalent electrical model, we see that the currents (and therefore the cell voltages) through each cell progressively decreases from either end of the stack towards the middle. i.e. the middle cell could be seeing currents significantly lower than what the end cells in the stack are seeing.

This offset from ideal cell performance has three key impacts:

• The aging of all the component cells in the stack are not identical due to different current densities being experienced. Since all the cells are connected in series, the performance of the whole stack is strongly driven by the performance and the degradation of the worst performing stack.
• Although a certain fraction of nominal current density might be within acceptable limits for gas-crossover, those limits are only true for the end cells. Since the middle cells can operate at significantly lower current densities, this can cause formation of explosive gas mixtures at the middle cells. This further requires a margin on the lowest operable loads over and above what is required from evaluating diffusive and convective mixing of gases.
• At low enough current densities, the middle cells may cease to see any current, and become reverse polarized. This can have catastrophic consequences for the electrocatalyst which can be permanently damaged in those cells.

Therefore, parasitic losses pose a significant limitation on the lowest loads an electrolyzer can be safely operated in.

> Degradation

Longevity of electrolyzer stacks are determined by their Beginning of Life (BOL) voltages, and End of Life (EOL) voltages. Typically, all electrical systems are designed to handle the whole range. Modern electrolyzers are always operated in Current Controlled mode (CC), i.e. the load demand is translated into a current demand.

If Load demand is 50%, the current demand in the rectifier would be set to 50% of nominal current. The voltage is a function of the cell performance, and is lowest at the beginning of life (highest performance), and highest at the end of life (lowest performance). EOL is usually limited to +10% of BOL. If an electrolyzer is designed to operate at 2 V / cell, while the actual voltage that is being seen is 2.2V / cell, the stack is said to have achieved its end of life.

This degradation is normally quantified in uV/h (micro-volts / hour), or in %/1000h (% of BOL per 1000 hours). Typical PEM electrolyzers degrade at the rate of 5uV/h (0.25% / 1000 h) leading to a projected cell operational life of 40,000 hours. Modern stacks claim to have even lower degradation rates.

However, it is important to note that these degradation rates are considering continuous operation at reasonable stable load conditions.

Operation at intermittent loads, and daily shut-down and start-up actually have shown to cause faster degradation of the cells. Therefore, it might not be recommended to do frequent start-up and shut-downs. This is a typical challenge when it comes to operating the systems from intermittent renewable power.

System Losses

Some system component's power consumption does not scale down in relation to the actual electrolyzer load, thereby leading to lower system efficiencies at low load conditions.

Even at low loads, the balance of plant equipment needs to keep operating at near nominal aux power consumption. While aux power consumption is typically in the range of 4-5% of the total system consumption for a 1MW system, that is true only when the system is running at 1MW. Operating at lower loads requires a significantly higher percentage of energy being consumed by the auxiliary systems leading to lower system efficiencies.

Conclusions

Electrolyzers are limited by the lowest operating load they can support due to the possibility of formation of explosive gas mixtures, which poses a significant safety risk. This puts a constraint on the operation of electrolyzers, when they are deriving their power exclusively from intermittent renewable sources like wind and solar. Even at loads at which it can operate, but less than nominally, lower system efficiencies lead to increasing OPEX.

Higher degradation rates due to frequent ramp-up and ramp-down also has an impact on the lifetime of the stack.

To enable sustainable integration of hydrogen with renewables, it is therefore necessary to address these challenges. Multiple technologies are being developed which strive to address these specific issues, which we would address in a future article.

Author: Deeparnak Bhowmick, Principal Engineer - Control & Automation.

Editor: Karnika K, Senior Associate - Founder's Office.