IoT Energy Storage Systems

Energy storage systems (ESS) are one of the energy sources for IoT devices. An energy storage system is a system that is designed to capture or receive energy from the energy source(s), convert it into a form that the system can conveniently store and then convert it into other usable forms of energy at a later time when the need arises. In the case of IoT devices, the electrical energy from energy harvesting systems incorporated in the IoT devices is converted to storable forms of energy (e.g., electrostatic, electrochemical, chemical, etc.) and later converted into electrical energy to power the IoT devices when needed.

In some deployment scenarios, energy is stored in an energy storage system (e.g., a battery) and then drawn to power the IoT devices. In this deployment type, when all the energy stored in the battery is depleted, the battery must be recharged or replaced; otherwise, the device will be shut down. The device's lifetime is the time from when the device is deployed to when all energy stored in its energy storage system is depleted. The energy storage system should be sized in such a way as to maximise the lifetime of the device to minimise the maintenance frequency and cost. Increasing the energy capacity of the device may result in an increase in size and price, which may be undesirable.

A possible way to increase the device's lifetime without a significant increase in cost and size is to incorporate energy harvesters to harvest energy, which is used to supply the IoT devices and store any surplus in the energy storage systems for later use. When the device is in sleep mode, most energy harvested is used to charge the energy storage system. In this case, the energy storage system can be a battery, a capacitor/supercapacitor/ultracapacitor, or a hybrid (a combination of more than one energy storage system to exploit the benefits of each of them). In this case, the energy storage system is designed to ensure a rational balance between the energy harvesting, storage, and consumption processes and maximise the devices' lifetime.

One of the responsibilities of IoT system designers and developers is to choose appropriate energy storage systems. The choice of the energy storage system will depend on the design goals, technical constraints, and other business criteria. Some of the design requirements to be considered when designing an energy storage system for IoT include the following:

• Safety, convenience, durability (durable operations) - the energy storage system is safe (less likely to explode or become flammable). It should have a more prolonged health (should operate for long without requiring replacement).
• Energy density - Energy storage systems (ESS) with higher energy densities can store more energy per unit of mass or volume, reducing the cost, size, and weight of IoT devices, which also facilitate mobility.
• Charging speeds - Energy storage systems with fast charging speeds are preferable.
• Ability to charge the ESS with small currents since the energy harvested from IoT energy harvesters is minimal.
• Ability to deal with peak power demand - the ESS should handle peak load demand, especially during peak communication or computing load demand.
• Long-term storage - the ESS should be able to store the energy for long enough to ensure that it can power the device if the energy-generating source is absent for some time.
• Cycle life - the ESS should have a large charge/discharge cycle to ensure longer cycle life and less need to frequently maintain or replace the ESS like the case with batteries.
• Cost - ESS made from elements or minerals abundant in nature are preferable as they will be cheaper. Most batteries are made from lithium, a relatively expensive mineral compared to sodium, which is very abundant in nature. Efforts are being made to produce solid-state batteries from sodium, which may eventually lead to cheaper batteries when technologies to have these kinds of batteries mature.
• Mobility - The batteries should be lighter to facilitate mobility.
• Size - In some IoT applications, especially in smart health care, it is desirable to ensure that the device's size is as small as possible, and a small-sized ESS is required.
• Environmental sustainability - choosing the ESS in such a way as to maximise the cycle life minimises the frequency of replacing the ESS, which ensures environmental sustainability. The ESS could also be manufactured using materials that are easily disposed of.
• Scalability - choosing durable ESS ensure scalability of IoT deployments as the limitation to scalable IoT deployments is dealing with ESS-related maintenance issues.
• Little or no energy leakage - energy leakage is a significant problem, and the ESS chosen should not have high energy leakage.

IoT devices can be powered with rechargeable and non-rechargeable batteries. The first requires a charger circuit (built-in or as an external device), while the second is suitable for ultra-low-power devices that can operate on a single battery for a very long time. Devices with non-rechargeable batteries allow the user to replace the battery (mechanically); that is not always the case for IoT devices powered with rechargeable ones.

Non-rechargeable batteries are available in standard sizes such as AA, AAA, C, and D and coin-size ones such as LR44 or CR2032.

Rechargeable batteries include transient technologies such as Nickel-Cadmium batteries (NiCd) and Nickel-Metal Hydride batteries (NiMH), which were modern in the 1990s and the beginning of the 21st century. They are replaced with Lithium-ion (LiIon) and Lithium-Polymer (LiPo), which present higher reliability, lack of memory effect and higher energy density. Still, they are also much more demanding on battery maintenance, including charging, discharging, operation temperature monitoring, and storage.
Lead-acid batteries are still common, but their application in IoT is limited due to their size and weight (low energy density), so they usually work as backups, e.g. in the context of green energy storage.

LiPo
Lithium polymer battery is a subtype of lithium ions. A single cell of the Lithium Polymer battery is usually in the form of a flat cuboid (figure 1). The single cell's standard reference voltage is 3.7V. When fully charged, the cell reaches 4.2V and should never be charged over this limit. On the other hand, LiPo cells cannot be discharged below 3.3V (some to 3.0V). The discharge curve is predictable and common for both LiPo and LiIon. A sample discharge curve is present in the figure 4. Thanks to it, it is possible to estimate the remaining energy.
Single-cell voltage is low for most applications, so serial-connected cell stacks (battery packs) are used. Serial connections used to be referenced with S (capital letter s). So, e.g. 3S represents a battery composed of 3 cells, connected in serial. In the case of the serial connection, the battery's voltage is the sum of each cell. Thus, 3S represents a battery of 3*3.7V=11.1V (reference) or 12.6V when fully charged and 9.9V when fully discharged.
In the case of charging the serial-connected battery packs, charging requires a separate balancing of each cell and usually requires a so-called microprocessor charger. RAW battery packs composed of more than 1 cell have two terminals: main and auxiliary for load balancing (figure 2). The connection for charging the sample 5S battery is present in figure 3.

LiIon
Lithium Ion batteries are widely used in electronic equipment nowadays. Their physical form is similar to LiPo ones, but there are also cylindrical units. Similarly to LiPo, LiIon single cell is nominal 3.6V or 3.7V. Charging and discharging require advanced control, and all warnings mentioned above regarding charging, discharging and maintenance of the LiPo cells also apply to LiIon.
The popular model for LiIon cell is the 18650 (figure 5) - the number comes from its dimension: a cylinder 18mm wide and 65mm high. Typical capacity is 2000-2500mAh per single 18650 cell.

Besides the 18650, other sizes are available, such as 14500 (similar to AA size battery) with a capacity of hundreds of mAh or 26650 with a capacity exceeding 10000mAh, designated for high-rate applications such as actuators.

Critical applications, such as use in electric cars, involve advanced cooling and heating systems to ensure optimal battery pack performance and charging conditions.

Most rechargeable batteries require a Battery Management System (BMS) that controls the charge and discharge of the RAW cells. It is essential, particularly in the case of the Lithium Polymer batteries and Lithium Ion ones.
BMS prevents overcharging and over-discharging and sometimes controls battery temperature, limiting charging current if the battery is overheating. Its general purpose is to keep the battery in good condition for a long time and prevent battery damage.

Raw LiIon and LiPo cells are commonly available, and there are also protection and charging modules in the form of electronic PCBs for self-assembly, e.g., a dedicated module for an 18650 LiIon cell as in figure 6 and figure 7. BMS can also be integral to the IoT device's power module, e.g. figure 8. Those boards usually contain DC-DC converters, providing a stable voltage of 3.3V and/or 5V for powering IoT devices.

Alternative energy storage systems that can be deployed to compensate for the limitations of batteries are capacitors, supercapacitors, or ultracapacitors. They can be deployed alongside batteries (hybrid deployment) and may eventually replace batteries in some IoT deployments. Some of the limitations of batteries that can be resolved with the use of capacitors, supercapacitors, or ultracapacitors are

• Limited cycle life - the limited cycle life requires that batteries should be replaced frequently, resulting in multiple challenges such as high and tedious maintenance costs (as it is difficult to service a vast number of IoT devices to replace or charge the batteries), degradation of the quality of service (as devices can be shut down when all the energy stored in batteries is depleted), and challenges in disposing of batteries (as vast amounts of batteries are required to be disposed of).
• Inability to handle peak power load demand - Small batteries are often not able to handle peak power load demands (which may result from peak communication or computing loads), which will require that the battery should be discharged at a higher C rate, which may be unhealthy or detrimental to the battery.
• Slow charging and discharging process — Batteries' charging and discharging speeds are relatively slow compared to those of capacitors, supercapacitors, and ultracapacitors.
• Charging and discharge inefficiencies - The magnitude of the energy harvested from the ambient environment or external sources using the small energy harvesters in IoT devices is very small (in the order of a few hundred microwatts or milliwatts) to charge batteries but can effectively charge capacitors, supercapacitors, and ultracapacitors due to their high charging and discharging efficiencies.
• Sustainability challenges — Since batteries may be replaced regularly due to their short lifetime, there is a growing challenge of disposing of battery waste without causing significant environmental damage. Some materials used to make batteries are toxic to the environment, and frequent disposal of large amounts of batteries poses a potential environmental risk.

There is an increase in the adoption of capacitors, supercapacitors, or ultracapacitors as alternative energy storage systems in IoT devices due to their advantages as alternative energy storage systems. Some of their advantages include:

• Longer cycle life — The cycle life of capacitors, supercapacitors, or ultracapacitors is far greater than that of batteries. So, there is no need to frequently replace them, reducing maintenance costs and e-waste generated by frequently changing batteries. Supercapacitors can reach up to one million charge/discharge cycles, eliminating the limited cycle life problems often experienced when using batteries as energy storage systems.
• High power densities — The high power densities make it possible to charge them with small currents (since the amount of power produced by IoT energy harvesters is very small) and also handle peak power load demands (which require the delivery of relatively large power to the IoT devices).
• Sustainability - Since there is no need to change the energy storage systems frequently, the amount of waste produced is relatively small. The supercapacitors are also made from materials that can be easily recycled.
• Faster charging and discharging speeds - Capacitors, supercapacitors, and ultracapacitors can be charged relatively fast compared to batteries.

Although using capacitors, supercapacitors, and ultracapacitors has many advantages compared to batteries, they also have limitations.

• Inability to store energy for long — One limitation of this type of energy storage system is that it does not keep power for long, resulting in the short lifetime of the IoT devices (the time required to deplete all the energy stored in the capacitors, supercapacitors, and ultracapacitors).
• Size and cost limitations — One possible solution to the problem of short device lifetime resulting from the quick discharge of capacitors, supercapacitors, and ultracapacitors is to increase the energy storage capacity. However, this will increase the size and cost of the IoT devices, which is not desirable in most IoT applications, as devices are required to be as small as possible.
• Decrease in energy capacity - When a supercapacitor reaches the end of its life, its energy capacity may drop to about 70% of its original value, limiting its ability to meet the energy storage needs of IoT devices.
• Energy losses — They suffer from energy losses resulting from internal energy distribution and current leakage, which result in the wastage of the energy harvested and stored. Power leakage leads to low utilization of the harvested energy, and a portion of the harvested energy leaks away instead of powering the IoT devices.

Unlike batteries, supercapacitors have a lower energy density but do not suffer from cyclic degradation, similar to the case with battery cells. Ceramic capacitors are often used as an energy storage system to store energy harvested by energy harvesters incorporated into IoT devices because of their low degradation, low current leakage, and expected increase in energy densities. Thus, they are an excellent candidate to be adopted as energy storage systems deployed in industrial IoT devices ^[1] and IoT devices in other sectors.

Although electrochemical energy storage systems (e.g., batteries) and electrostatic energy storage systems (e.g., capacitors, supercapacitors, and ultracapacitors) are the most popular energy storage systems used to store energy to be used to power IoT devices, other energy storage systems can be used, especially at the transport layer (internet access and core networks) and fog and cloud computing layer. Some of these other energy storage systems may not be convenient for IoT devices. Some of them include the following:

• Chemical energy storage systems - chemical energy storage systems convert the electrical energy delivered to them into chemical energy, which can then be converted into electrical power to supply the IoT systems later. One popular example of a chemical energy storage system is the hydrogen energy storage system. In a hydrogen energy storage system, electrical energy is converted into hydrogen, which is then stored. One of the approaches often used to produce hydrogen is water electrolysis, which produces hydrogen and oxygen. The hydrogen is then stored and later used as fuel in a fuel cell to generate electricity to power the IoT infrastructure (e.g., base stations and data centres). Compared to battery energy storage systems and supercapacitors, battery energy storage systems are inefficient as much energy is wasted. However, much research is being conducted by major energy and car companies and academic institutions to improve the efficiency of hydrogen energy storage systems, as it is expected that hydrogen energy storage systems should be among the top innovative technologies for future green economies.
• Mechanical energy storage system - mechanical energy storage systems can convert electrical energy into mechanical energy (potential or kinetic energy), which can then be converted into electrical energy to power IoT systems later. The most popular mechanical energy storage systems include pumped hydro, flywheels, and gravity energy storage systems. Mechanical energy storage systems are simple to design, as this technology has existed for hundreds of years. One of the limitations is that they have very low energy density and are also very inefficient.

The various energy storage systems that we have discussed above have their advantages and drawbacks. One possible way to exploit the advantage of some energy storage systems and eliminate the limitations imposed by some energy storage systems is to deploy more than one energy storage system. An energy storage system that consists of more than one energy storage system is called a hybrid energy storage system. The deployment of hybrid energy storage systems (more than one energy storage system) improves the overall performance of the energy storage system in terms of energy density, reliability, and the cycle life (or lifespan) of the energy storage system. It also reduces the overall cost of the energy storage system.

In IoT devices, batteries and supercapacitors can be deployed as a hybrid energy storage system. The advantage of supercapacitors is that they can be charged faster, even with small currents (typical of the currents delivered by IoT energy harvesting systems). The supercapacitor can also handle peak power loads and have a longer cycle life than batteries. The limitation of supercapacitors is that they cannot store energy for a long time, but batteries can store energy for a long time. Therefore, a battery and supercapacitor can be installed in an IoT device to provide a hybrid energy storage system that takes care of the limitations of both and exploits their advantages to offer improved performance.

Batteries are often used as an energy storage system in base stations and cloud data centre sites powered by renewable energy. Due to the limitation of cycle life and power density, a hybrid energy storage configuration consisting of a supercapacitor and battery can be considered. Another kind of configuration is a battery and a hydrogen energy storage system. When the battery is full, any additional energy harvested is lost (in some installations, it is passed through a dumb load to dissipate it).

In a battery-hydrogen hybrid energy storage system configuration, when the battery is full, the additional electrical energy harvested is used in water electrolysis to split water molecules to produce hydrogen and oxygen. The oxygen is then stored and later used as fuel in a fuel cell to generate electricity when necessary. This type of hybrid energy storage is beneficial for seasonal energy storage where during the season when the conditions for energy harvesting are favourable (e.g., during summer), a lot of energy is harvested and stored in the form of hydrogen, which is then used during winter to generate electricity when the energy harvesting is not enough.