Understanding Batteries
Profiting from pairs of redox reactions that generate electric current.
Batteries are better living through chemistry. Improved battery technology is a key enabler in the transition from fossil fuel to clean energy, making solar power dispatchable at all scales, from individual devices to grid-stabilizing solar farms.
Read this article for a deep dive into battery technology. Knowing more about how batteries work can help you buy better batteries, take better care of the batteries you already own, or make smarter investments in companies that produce or use batteries.
I wrote this article because much of the information on the Internet about batteries is out-dated, confusing, contradictory, and often wrong. As I’ve been researching for my handbook on sustainability, I’m writing up and publishing articles like this to organize my notes and consolidate my understanding of complicated topics. As the old adage goes, the best way to learn a subject is to teach it! It’s helpful to me (and I hope to you, too) to have reference articles that are accurate, clear, and thoroughly researched.
Read on if you’re interested in taking a deep dive into what’s happening inside each battery you own, where battery technology is heading, and what that will mean for how we will all live after 2030 or so.
Batteries and Other Energy Storage Technologies
Benjamin Franklin coined the term “battery” for an assembly of Leyden jars, arranged together to increase their power in the same way a group of cannons can be used in military campaigns. Franklin noticed that one side of each jar held a positive charge, and the other a negative charge. Connecting the two sides allows an electrical current to flow.
Today we would call Leyden jars capacitors, not batteries, since they store charge in an electric field. In modern parlance, batteries refer to cells that store energy in chemical bonds. Both capacitors and batteries allow energy to be stored and then released as electricity. In general, capacitors can store smaller amounts of energy and release it quicker than chemical batteries. Technologies like pumped hydroelectric power use gravity to store larger amounts of energy than batteries.
Fuel is another way to store energy in chemical bonds. Thermal power plants release the chemical potential of fuel as heat to expand fluids to spin turbines to generate electricity. Compared to power plants, batteries are a much cleaner, simpler and more efficient way to generate electricity from chemicals.
Batteries allow electricity to be stored whenever and wherever it can be generated conveniently, then used whenever needed. Like solar power, batteries are relatively easy to scale from tiny to enormous.
Since fossil fuel is the primary energy storage technology for the world economy, that provides a baseline for evaluating battery technologies.
The gravimetric energy density (also called “specific energy”) of fossil fuel varies between 23.9 megajoules (MJ) per kilogram (kg) for bituminous coal to 55.5 MJ/kg for natural gas. On the basis of specific energy, all batteries available in commercial quantities in 2023 can be considered “very bad.” The specific energy for lithium ion batteries developed prior to 2020 varies between 0.36 and 0.954 MJ/kg.
Despite the severe limitations of the current generation of batteries, an industry group predicts that demand for them will increase significantly between 2023 and 2030 for both grid electricity storage and for electric vehicles. So there is a race on to build a better battery. Since 2021, more than $13 billion per year is being invested in battery materials, manufacturers and recyclers. While batteries may not surpass fossil fuel in energy density by 2035, they are closing the gap.
Battery Science
To understand how batteries work, it’s helpful to know a little chemistry and electricity — especially a few technical terms and core concepts.
The smallest piece of anything is called an atom. Every atom has a nucleus, which has a positive electrical charge, and an electron cloud where electrons can be. Electrons have a negative charge. Depending on how many electrons are in the clouds around the nuclei of a group of atoms, that group can be a positive ion (cation), a negative ion (anion), or neutral with no net electrical charge. Electrical charge is an integer value, so a cation might have a +1 or larger charge, and an anion might have a -1 or smaller charge.
Groups of atoms interact in reactions that rearrange their electron clouds. A reduction reaction makes the group of atoms have a more negative charge, i.e. you reduce the charge level. You might reduce a +2 cation to a +1 cation, or reduce a +1 cation to neutral (0 charge). An oxidation reaction makes the group of atoms have a more positive charge. You might oxidize a -2 anion to a -1 anion, or oxidize a -1 anion to neutral (0 charge).
Reduction and oxidation reactions that occur in pairs to transfer electrons from one group of atoms (“chemical species”) to another are called redox reactions. On one side of a battery you might have a reduction reaction that is reducing cations from +1 to 0, and on the other side have an oxidation reaction that is oxidizing anions from -1 to 0. The reduction reaction gains electrons lost by the oxidation reaction.
Batteries operate due to redox reactions that transfer electrons from one side of the battery to the other. However, rather than simply transferring electrons directly from one chemical species to another, a battery is cleverly designed so that electrons must travel through an external electrical circuit first. The charge that propels the electrons along this circuitous path can be harnessed to do work.
Electricity is the flow of electric charge. Positive charge flow is called “current,” while electrons flow in the direction opposite to current. In any direct current circuit, in which current flows in one direction rather than alternating directions, the terminal called “positive” is usually the source of current and the terminal called “negative” is usually the source of electrons.
Current always flows from an area of higher electrical potential to lower electrical potential. The difference in potential is called voltage. If you use a voltmeter to measure the difference in electrical potential between the positive and negative terminal of a battery, you would expect to see a positive number. A negative number would mean that for some reason current is being pushed out of the negative terminal. Or you might just have your wires crossed.
How Batteries Work
All battery cells consist of three essential components:
- A positive electrode.
- An electrolyte.
- A negative electrode.
The positive electrode contains materials for a spontaneous reduction reaction, using up any free electrons that happen by. The other electrode, the negative electrode, has materials for an oxidation reaction, throwing out free electrons. An electrolyte separates the two electrodes electrically but connects them ionically.
Providing an electrical conductor between electrodes completes a circuit that allows a pair of redox reactions to happen, sending electrical current outside the battery from positive to negative electrode and electrons in the opposite direction while sending ionic current inside the battery from negative to positive electrode.
Inside the battery cell, the electrolyte allows charged ions but not electrons to flow. Any anions (negatively charged ions with extra electrons) in the electrolyte flow from the positive electrode to the negative electrode where they are used in the oxidation reaction, giving up a free electron which then travels outside the battery through the electrical circuit. Any cations (positively charged ions lacking electrons) flow from the negative electrode to the positive electrode where they are used in the reduction reaction, gaining electrons that arrive from the electrical circuit.
When a battery cell is releasing energy by allowing a pair of spontaneous redox reactions to occur, it is acting as a “galvanic” or “voltaic” cell. The chemical reactions are producing a voltage difference between positive and negative electrodes. The electrode where the reduction reaction occurs is called a “cathode.” Cations flow toward, electrons flow in, and current flows out of a cathode. The electrode with the oxidation reaction is called an “anode.” Anions flow toward, electrons flow out, and current flows into an anode.
A spontaneous chemical reaction happens unless you block it. If the positive and negative electrodes are not connected, electrons can’t flow because the electrolyte blocks them. That prevents the redox pair from happening: the cathode needs more electrons, and the anode needs fewer. When the electrical circuit is connected, the cathode can gain electrons and the anode can lose electrons. That sends cations and anions shuttling across the electrolyte to catch and release more electrons to send through the electrical circuit.
Some batteries have anions in their electrolyte, such as alkaline batteries which use hydroxyl anions to shuttle extra electrons from the positive to the negative electrode. Other batteries have cations in their electrolyte, such as lithium ion batteries, which use lithium cations to shuttle electron holes to the positive from the negative electrode. It’s possible to have batteries with both anions and cations, which would be like ships passing in the night shuttling charge across the electrolyte ocean.
One way to think of it as the ions in an alkaline batteries pushing electrons through the circuit from the negative electrode, whereas the ions in a lithium ion battery are pulling electrons through the circuit from the positive electrode. Either way, the flow of ions allows the pair of redox reactions to occur that keeps the electrons flowing. When the circuit is disconnected, the flow of electrons stops, the flow of ions stops, and the redox reactions have a hard time continuing.
One issue with batteries is that chemical reactions can happen even when the battery is not connected to an electrical circuit. This is called self discharge. Some battery chemistries are more prone to self discharge than others. The temperature of the battery can often speed up or slow down the rate of chemical reactions. The slower the rate of unwanted chemical reactions, the longer the battery can store a charge.
Another issue issue with batteries is what happens when all of the ions have moved from one electrode to the other. In “primary” batteries, it’s game over. These types of batteries are not rechargeable. Once all the ions have been used up in redox reactions, there’s no force moving electrons between electrodes. With non-rechargeable battery chemistries, there’s no way to get the ions to move back across the electrolyte.
But some redox reactions can be reversed. In rechargeable batteries, ions flow in one direction spontaneously, but can be forced to flow in the opposite direction by applying current to the electrodes. If you connect two different batteries together positive to positive and negative to negative, one battery’s positive terminal may have a higher electrical potential than the other battery’s. Current will flow from the higher potential (stronger battery with a higher voltage) to the lower potential (weaker battery with a lower voltage).
Think about what happens if you connect two cells positive to positive and then negative to negative. If the two batteries have the same electrical potential at either electrode, current can’t flow. But if one battery has a higher voltage than the other, once an electrical circuit is made, current will flow from the positive electrode of the stronger battery to the positive electrode of the weaker battery. Electrons flowing the other way will allow a reduction reaction to occur in the stronger cell’s positive electrode.
Let’s follow the path of an electron from the weaker cell. Normally electrons flow into the positive electrode. But wait, electrons are flowing out of the positive electrode of the weaker cell. What’s going on there?
The weaker cell becomes an “electrolytic” cell. Normally the chemicals in the positive electrode would undergo a reduction reaction that gains electrons, but with the stronger battery sucking electrons out, that favors an oxidation reaction to release electrons. With so many electrons getting pulled out of the positive electrode of the weaker cell, its natural tendency to reduce is overwhelmed. Instead, it starts to oxidize groups of atoms, making them more positively charged. That repels cations and attracts anions.
When the electrolytic cell’s positive electrode becomes an anode, oxidizing and releasing electrons, its negative electrode starts acting as a cathode, reducing reducing groups of atoms to be more negatively charged by gaining electrons. That attracts cations and repels anions.
In short, in an electrolytic cell, an externally supplied charge forces the flow of current and electrons to drive chemical reactions which wouldn’t spontaneously occur to move ions “backwards” between electrodes. To charge a battery, you must apply a voltage to the positive terminal that is high enough that will overcome the natural tendency of current to flow out. Different chemical reactions produce different electrical potentials, so the voltage of a cell and the voltage require to charge that cell depend on its chemistry.
What happens if you try to recharge a non-rechargeable battery? In the best case, nothing. The charge forcing current into the positive electrode prevents it from sending current backward out toward the charging source and the electrolyte prevents the current from flowing forward electrically inside the cell. The current can’t force ions across the electrolyte because the chemical reactions in the positive electrode don’t work in reverse (or don’t work very well in the reverse direction).
But current might start leaking forward through any paths it can find to flow electrically inside the battery pack and inside cells from positive to negative electrodes. If current can flow electrically (rather than ionically) inside the battery, it is likely to meet resistance, which heats up the electrodes and electrolyte. In the worst case, trying to recharge a non-rechargeable battery might cause it to explode if heat builds up fast enough inside a cell.
As an aside, note that an electrolytic cell could be used to produce ions or other products of any pair of redox reactions.
Electrolyzing water to produce hydrogen and oxygen gas is an example. Sending current through an electrolytic cell can split water molecules, drive hydrogen ions from positive to negative electrode, and combine hydrogen ions into hydrogen gas molecules.
Connecting, Stacking and Packing Cells
A battery (reminiscent of a regiment of artillery cannon) is an assembly of cells that are all connected together by electrical conductors in a way that allows current to flow from a positive terminal to a negative terminal on the pack that contains all the cells.
The voltage of each cell, the difference in electrical potential between its positive and negative electrodes, depends on its chemistry. For example, a nickel metal hydride cell is 1.2 V, an alkaline cell is 1.5 V, and a lithium iron phosphate cell is 3.2 V. Going back to Benjamin Franklin’s observations, you can stack several cells together in series to increase the voltage of the battery.
Connecting cells in series means connecting the positive electrode of one to the negative electrode of another. The positive terminal for the load is connected to the positive electrode of the first cell in the series and the negative load terminal is connected to the negative electrode of the last cell in the series. The voltage supplied to the load is the sum of all the cells in the series. For example, a 3S LiFePO4 battery, which contains three 3.2 V cells connected in series, supplies 9.6 V to the load (3.2 + 3.2 + 3.2).
When cells are connected in series, each one adds to the electrical potential. For each cell, the positive electrode has a higher potential than the negative electrode. The negative electrode starts at whatever potential the cell behind it in the series is at, and then the positive electrode electrode boosts that potential. It’s like climbing a ladder. Each cell starts at one rung and steps up to the next.
Current behaves differently. The amount of current flowing through the electrical conductor depends on how many electrons are being sent through. When cells are connected in series, each cell isn’t adding more electrons on the conductor. Every electron needs to hitch a ride on an ion to get through each cell. The electrons can climb higher on the ladder of electrical potential, but only as many as can find a ride on an ion through a cell’s electrolyte can get on the electrical conductor.
To send more electrons through the conductor, connect cells in parallel. A 3P LiFePO4 battery, for example, contains three lithium iron phosphate cells connected in parallel. The voltage of the battery is 3.2 V, but the amount of current the battery can supply is three times that of a single cell.
Many more electrons can move through the electrical circuit because they only have to go through one battery to make a round trip.
But the electrical potential of each electron is only raised as much as the difference between the negative and positive terminal of a single cell.
Note that if you have three cells, the total amount of energy that you can store and power that you can produce is the same regardless of how you stack and pack them. Power is measured in watts, which are volts times amps. For any cell chemistry, the volts of each individual cell are the same. Volts measure how high the cell can raise electrons in electrical potential. But the amps depend on the physical size of the cell. Amps measure how many electrons the cell can move at once.
If you have three lithium iron phosphate cells, they will each be 3.2 V no matter their physical size. If they are the same size, they will all be able to produce the same number of amps (i.e. move the same number of electrons at once). For example, you might have three cells that are rated 3.2V 10A (power rating) 9600Wh (energy rating). You could arrange them to make a 3S battery that can produce 96 watts of power for one hundred hours by stacking the cells in series to provide 9.6V and 10A. Or you could pack the cells in parallel to provide 3.2V and 30A with a 3P battery that produces 96 W of power for one hundred hours.
Measuring Battery Performance
The purpose of a battery is to store energy and then release it as electricity. When connected to a complete electrical circuit, batteries produce a direct current flow from positive to negative electrode. Batteries are rated primarily by
- capacity, both the total amount of energy they can store and release and how much energy per time (power) they can discharge, which is a combination of the voltage (potential difference between the starting and ending point of the current flow) and the amperage (the number of electrons flowing),
- recharge rate,
- allowed depth of discharge, and
- cycle life before capacity fade.
The voltage, or potential difference between positive and negative electrodes, of a battery starts high and falls as the redox reactions occur. This tendency means that a battery’s power output also starts high and then decreases. As rechargeable batteries go through discharge-recharge cycles, they are able to store less and less energy due to entropy. The materials in the electrodes become disordered and accumulate contaminants which interferes with the redox reactions.
Historically, battery capacity was specified in terms of amp-hours, which is a measure of the number of electrons that the battery can store at its negative electrode and send to its positive electrode. The number of electrons a battery can store in an electrode depends on the number of ions it can store and the number of electrons each ion can shuttle. A redox reaction pair that reduces by -2 and oxidizes by +2 means that each ion carries two electrons, so per ion a -2/+2 redox reaction battery can store twice as much energy as a battery that uses a -1/+1 redox reaction.
A unit called “C” is defined as the power that would, over an hour, discharge or recharge the complete capacity of a battery. A 1 C discharge rate means the battery’s spontaneous current flow can move all the electrons from the negative to the positive electrode in one hour. A 1 C recharge rate means applying external current can force all the electrons back from positive to negative in one hour.
Modern batteries are beginning to specify their energy capacity in terms of watt-hours in addition to (or instead of) amp-hours. When batteries were almost all 12V, it made sense to specify their energy in amp-hours. You could multiply the amp-hours by 12 to get the actual energy capacity. Now that voltages might vary, consumers don’t want to have to do the math themselves, so battery manufacturers are starting to list the watt-hours, which is voltage times amp-hours.
It’s similar to the way that the brightness of light bulbs used to be marketed by their power consumption (a 100 watt light bulb was much brighter than a 60 watt light bulb). But as LED light bulbs have appeared on the market, advertising wattage no longer made sense, because a 20-watt LED light bulb is much brighter than a 60-watt incandescent light bulb. So light bulb manufacturers are switching over to advertising the actual brightness, which is measured in lumens.
If a battery has an energy storage capacity of 100 kWh and a 1 C maximum discharge rate, that means that its power output is 100 kW. Producing 100 kW of power for one hour requires 100 kWh. If the battery had a 2 C maximum discharge rate, that means that in one hour the battery could discharge twice as much energy as it can store. So a 2 C rated battery with an energy storage capacity of 100 kWh could discharge 200 kWh per hour, which would be a power rating of 200 kW.
If you know the battery’s energy storage capacity and discharge and recharge C rate you can determine the following:
- How much power the battery can produce at peak ion flow rate.
- How long the battery can operate at peak power.
- How quickly the battery can recharge.
For example, let’s say a type of battery has a 2 C discharge rating and a 1 C recharge rating. You buy a 10 kWh size of this battery.
- This battery can produce 20 kW of power .
- This battery can operate for 30 minutes (half an hour) at peak power.
- This battery can recharge from fully depleted to fully charged in one hour.
If you bought a 50 kWh size of this same type rated as 2 C discharge and 1 C recharge, that battery could produce 100 kW of power for 30 minutes and recharge from empty to full in one hour.
Some battery chemistries like lead acid can hold more charge than is wise to discharge. In these chemistries, letting all the ions move from one electrode to the other is a bad idea because not all of them will be able to move back again. A common rule of thumb with lead acid batteries is to operate them so that they always keep half of their charge capacity in reserve. The allowed depth of discharge is 50%. This means that if the battery energy storage capacity is 100 kWh, you should only use 50 kWh of it before recharging. Only if you don’t care about ever recharging the battery can you let every last ion flow.
Other battery chemistries, like lithium iron phosphate, allow a 100% depth of discharge. In this type of battery, you can let all the lithium cations move from negative to positive electrode and have confidence that when you recharge the battery, you can herd all the cations to move back to the negative electrode. So if you have a LiFePO4 battery with an energy storage capacity of 100 kWh, you can use all 100 kWh before needing to recharge.
Over time, entropy happens. Discharging and recharging tends to accelerate entropy. This disorder adds up until most of the electrons are no longer able to find their way back and forth. The number of discharge-recharge cycles is a big concern for the affordability and practicality of products that use rechargeable batteries. Imagine if you bought an electric car that only lasted 10,000 miles before its batteries needed to be replaced.
Lead acid batteries have a cycle life between 1,000 and 1,200. Lithium ion batteries have a cycle life between 3,000 and 5,000. Iron flow batteries have a useful life in excess of 20,000 cycles.
Confusing Terminology
The way stock analysts talk about “cathode” and “anode” materials can be confusing.
In every type of battery cell, you always attach the positive wire to the positive electrode and the negative wire to the negative electrode. But which electrode is acting as a cathode and which as an anode depends on whether the cell is producing or accepting current.
According to Phil Karn on Quora:
The standard convention is that the electrons enter a device through its cathode and leave through its anode. For a device like an electron tube that absorbs power, this means the anode is positive with respect to the cathode. For a primary battery that generates power, the anode is negative with respect to the cathode.
A rechargeable battery can either absorb or produce power, so technically its anode and cathodes change roles. Here the convention is to consider only the discharge case, so the anode is always the negative terminal and the cathode is always the positive terminal — the reverse of diodes and electron tubes.
And it gets even more confusing when you remember that “conventional current flow” is defined to be from positive to negative, opposite the actual flow of electrons. We can blame Benjamin Franklin for getting this one wrong, but it’s too late to change now.
Stock analysts who write about companies that are working on materials that can be used as electrodes don’t call them “positive” or “negative” electrodes, instead they call them the “cathode” and the “anode.” A good material for a positive electrode in a rechargeable battery will spontaneously react to reduce its electrical charge (acting as a cathode), but be able to oxidize when current is sent into it (acting as an anode). It’s usually the case that when a stock analyst talks about a “cathode” material they mean a material that can be used in the positive electrode of a galvanic cell.
A second area of confusion is between self-contained batteries and electricity storage systems that allow an external source of chemicals to be supplied to their electrodes for the redox reactions.
Self-contained batteries include all of the material necessary for both redox reactions, and those chemicals remain inside the battery for its entire useful life. This has advantages and disadvantages. One advantage is that self-contained rechargeable batteries can be used in a wide variety of environmental conditions and require nothing more than a periodic supply of electrical current at a voltage sufficient to recharge them. A second advantage is that the battery does not change in weight and does not emit anything (other than hydrogen gas in some circumstances). A third advantage is that materials in the battery can be recycled.
A disadvantage of a self-contained battery is that the capacity of the battery is fixed. A second disadvantage is that the battery needs to be bigger and heavier than it could be if materials could pass through it, especially if gases could be absorbed from and released to ambient air.
Electricity storage systems that are not self-contained, but which can take in and release chemicals necessary for the reactions that generate electricity, are called fuel cells, air batteries, and flow batteries. Since materials can be added to their positive and negative electrodes, these systems can generate electricity as long as the necessary chemicals are supplied.
Consider how a fuel cell could use hydrogen and oxygen to electrolyze water to produce hydrogen and oxygen. (Doing this has no practical value, but simply demonstrates how a fuel cell and an electrolyzer run the same reactions in opposite directions.)
A fuel cell is a galvanic cell which requires oxygen gas in its positive electrode, an electrolyte that allows H+ cations to move from negative to positive electrode, and hydrogen gas in its negative electrode. Hydrogen gas could be supplied to the negative electrode from a gas tank, and oxygen gas could be obtained from the atmosphere (which is about 18% oxygen gas by dry volume).
The same pair of redox reactions could be designed to occur in a self-contained battery with a fixed amount of hydrogen and oxygen, but a fuel cell is designed to take in both oxygen and hydrogen gas from an external source and to emit water as a waste product. Using external sources of oxygen and hydrogen provides more flexibility. The same fuel cell can run a shorter or longer amount of time, depending on how much hydrogen and oxygen is supplied.
An electrolyzer is designed to take in water and produce both oxygen and hydrogen gas. The fuel cell and electrolyzer system could be self-contained. For example, you could use it to inflate or deflate a balloon. Run the system in one direction to converts water into hydrogen and oxygen gas to inflate the balloon; reverse direction to combine hydrogen and oxygen back into water. But a self-contained system could only make a fixed amount of gas from a fixed amount of water. Making the electrolyzer open to an external source of water allows you to continue making oxygen and hydrogen as long as your energy and water supplies last.
Similar to a fuel cell, a flow battery is supplied with chemicals for the redox reactions.
One chemistry for iron flow batteries uses iron chlorides and metallic iron, stored in external tanks in a liquid solution, which can be pumped through the electrodes. An advantage to this type of battery is that the capacity depends on the tank size, which can be scaled up independently of the size of the electrodes.
A final confusing bit of terminology are so-called “anode-less” batteries. Cells in these batteries, like every other battery cell, actually do have both a positive and a negative electrode. During manufacturing, however, the positive electrode material contains all of the cations and the negative electrode is empty. The first time the battery cell is charged, positive current forces cations from the positive electrode, across the electrolyte, to the negative electrode.
Reducing the amount of material used for the negative electrode reduces the cost of the battery and improves its gravimetric and volumetric energy density. You can store the same amount of energy using less stuff. In a state-of-the-art regular lithium ion battery circa 2023, the negative electrode contains carbon graphite, which acts as a docking station for lithium cations when the battery is fully charged. During discharge, the lithium cations migrate from the graphite across the liquid electrolyte toward the positive cathode.
An “anode-less” battery swaps the liquid electrolyte for a solid electrolyte. If the electrolyte is a solid, it can hold the lithium cations in place without the need for the graphite docking station structure in the negative electrode. It’s the battery equivalent of Erthos solar, which eliminates racking and puts solar panels right on the ground.
When an “anode-less” battery is fulled charged, the negative electrode typically is a metal, such as lithium, plated on the charge collector and held in place by the solid electrolyte to avoid shorting the cell. (A liquid electrolyte would be subject to metal bridges connecting the negative and positive electrodes, which is called a “short” because electrons can move a shorter distance than we want them to go.) During discharge, the negative electrode acts as an anode, oxidizing the lithium metal to send electrons through the charge collector and lithium cations across the electrolyte to the positive electrode, which is acting as a cathode. Only when an “anode-less” battery is fully discharged is the negative electrode empty again. Perhaps a better name for an “anode-less” battery would be a “disappearing anode” battery.
Building Batteries
At least 90% of the factories that will be building batteries in 2035 do not exist in 2023. This gives industry a chance to start over in terms of design and form factor. New batteries will be produced in sizes, shapes and performance capacities that are easy to manufacture on automated lines.
The experience of Panasonic and Tesla in their “Gigafactory” joint venture is informative. Between 2014 and 2023 they manufactured 7.3 billion battery cells without using any fossil fuel (thereby proving that fossil fuel is not required to assemble battery cells and packs). By comparison, an estimated 4 billion AA alkaline battery cells are produced each year by all manufacturers worldwide. Global battery production is expected to increase six-fold between 2019 and 2024.
Tesla started building battery packs for its cars using off-the-shelf battery cells in a standard size called “18650.” This refers to a cylinder 18 millimeters in diameter and 65.0 millimeters long. This is slightly larger than the standard AA size, which is 14.5 mm in diameter and 50.5 mm in length. When the joint venture began assembling batteries in the Gigafactory, they increased the size of each cell to 21 mm wide and 70 mm long. In 2020, Tesla announced that it would be producing batteries in a new size: 46 mm wide and 80 mm long.
The size and shape of a battery cell determines the amount of energy it can store and the rate of power it can provide. Each 4680 cell can store five times as much energy as a 2170 cell because it is physically larger and has more electrode material in it that can hold more ions.
Also in 2020, Tesla began using battery cells with lithium iron phosphate for the positive electrode in some products, eliminating cobalt from the formula. Until then, all their cells had used nickel cobalt aluminum and nickel cobalt manganese electrode materials.
The production lines in Tesla’s Gigafactory are optimized to manufacture cylindrical batteries. These cells are produced by coating the positive electrode material on aluminum foil and the negative electrode material on copper foil. Then the foils with a non-conducting separator membrane between them are wound into a cylindrical shape. Next they are inserted into a case, filled with liquid electrolyte, and sealed. Finally, the cells are subjected to charge/discharge cycles that form an electrochemical solid electrolyte interphase at the electrode surface, mainly on the negative electrode.
Many improvements are possible in the manufacturing process of cylindrical batteries that will reduce costs and increase production speed. For example, conventional practice is to use a wet process to coat foil with electrode material. This requires baking the electrodes to dry them out. A dry process to coat material on the foil would eliminate the time and energy required to bake electrodes. Another improvement would be a solid electrolyte. This would eliminate the need to fill the cases with liquid electrolyte. A solid electrolyte material would replace the non-conducting separator membrane between the electrodes and allow for so-called “solid-state anode-less” batteries in which the negative electrode would be created inside each cell during the charge/discharge manufacturing step.
Other companies are building batteries as pouches rather than cylinders. Manufacturing pouches requires coating foils with electrode materials, but rather than winding the electrodes and putting them into a cylindrical case, the electrodes can be stacked flat in a rectangular housing.
For self-contained battery technologies, materials that can be drop-in replacements for the electrodes, separator, or electrolyte in a cylindrical manufacturing line can get to market the quickest. For example, “Prussian white” (sodium, iron, carbon, and nitrogen) or lithium, magnesium, zinc, aluminum and phosphate can be used for a positive electrode material instead of lithium iron phosphate. For the negative electrode material, most lithium-ion cells use graphite or another form of carbon, but some companies are attempting to commercialize silicon electrodes.
Manufacturing techniques for the other class of battery technologies, types that can accept external materials supplied to their electrodes, are not as advanced as they are for self-contained batteries. Fuel cells, air batteries and flow batteries are being sold in much lower volumes. Many different approaches are still being validated in field trials.
In my view, fuel cells and flow batteries have had difficulty competing in portable or transportation applications due to the complexity of delivering materials to the electrodes. I believe air batteries are much more likely to be successful, since they can simply “breathe” ambient air as a source of oxygen. However, air batteries may require different manufacturing process than self-contained cylindrical batteries. Most cylindrical cells are sealed to prevent moisture or air from disrupting the chemical reactions inside them. Air batteries will need to enable air to flow only where it’s wanted and not where it will foul the reactions.
With the opportunity to build out a new battery industry many times larger than the incumbent industry comes the chance to standardize on a different voltage. Since the 1950s, batteries built using six lead acid cells providing 12 volts have been the standard for DC circuits, such as the starter and accessories in cars. This 12 volt standard is likely to be replaced by a 48 volt standard by 2035.
The higher the voltage, the smaller the wire needed to provide the same amount of power. Anything below 50 volts is considered relatively safe by the US Occupational Safety and Health Administration. For this reason, 48 volts, rather than a higher voltage is likely to become the new standard. For example, emerging standards like “Power over Ethernet” use 48 V. There’s no reason to stop at 24 V or 36 V; maximum cost savings and minimal regulatory hassle are available at 48 V.
Modern battery systems include power electronics between the electrical terminals and the pack of cells to ensure that the cells deliver and accept the correct amount of voltage and current to maximize their useful life. This allows manufacturers to substitute different cell chemistries, and even mix chemistries inside a battery system, while meeting performance standards in terms of the voltage and current the battery pack can provide.
These new battery management systems are possible due to advances in computer chips and information technology. Further improvements in these management systems will be possible as software begins to play a more prominent role than hardware in system architectures. In the past, systems have been hard coded and could not be modified after shipment to customers, but in the future, virtually all battery management systems will allow for network connections to provide for software updates to improve functionality over the lifetime of the energy storage system. These smarts and network connections will enable a new way to buy and sell electricity.
Using Battery Science to Make Smarter Investments
Science is a method of predicting the future with greater accuracy. Understanding battery science helps predict which companies will benefit as the world transitions from fossil fuel to clean energy.
Here are a few of my predictions of what will occur by 2035. If I’m correct, these will give you an advantage over other investors.
Prediction #1: The Energy Density of Batteries Will At Least Quadruple
Self-contained batteries are much heavier than air batteries. An air battery can use oxygen from ambient air in its positive electrode, greatly reducing the mass the battery contains. In February, 2023, Science magazine published the results of researchers at Argonne National Laboratory who have demonstrated a practical rechargeable lithium-air battery which they anticipate will achieve an energy density four times better than lithium-ion batteries based on lithium nickel chemistries.
Improving the energy density of batteries is a priority for batteries to compete with fossil fuel, which has an energy density two orders of magnitude better than the lithium-ion batteries available in 2023. Marginal improvements in energy density (i.e., anything less than doubling it) are insufficient for many applications such as air taxis and long-distance trucking, so researchers and companies have an incentive to make much better batteries that have higher energy densities than are possible with nickel-based lithium-ion batteries.
Prediction #2: The Cost of Batteries Will Decline
Chemistries that require expensive materials like cobalt, nickel and vanadium will be supplanted by chemistries that use affordable materials like carbon, sodium, iron, phosphorus, and zinc. With tens of billions of dollars just beginning to be invested in the battery race and new artificial intelligence tools coming online to help with basic materials science, our understanding of the possible pairs of redox reactions will greatly improve. This will allow engineers to design out toxic or expensive chemical species from future batteries.
Economies of scale will improve as manufactures invest almost a trillion dollars in the transition to electric vehicles by 2035. Better selection of low-cost raw materials to make batteries will be compounded by improvements in manufacturing. The raw material cost for batteries will decline and the time to fashion that material into battery cells will decline.
Prediction #3: The Reliability of Batteries Will Improve
Traditionally, batteries have consisted of little more than the bare essentials: two electrodes separated by an electrolyte. The chemistry of the cells and the physical connection between them have determined the amount of current and voltage the batteries produce.
Advances in modern electronics and computing allow battery makers to add more sophistication. Battery cells now are bundled with battery management systems which control the flow of current in and out of the individual cells. Battery packs can monitor their own temperature and maintain optimal conditions to maximize the useful life of each cell.
Prediction #4: The Demand for Batteries Will Increase
Batteries are already used for portable electronics, power tools, vehicles, emergency power back up, and grid-scale energy storage. As battery performance improves and costs decline, the demand for batteries will grow. Government mandates to reduce pollution and incentives to electrify will give a small boost to demand, but the biggest driver will be the superiority of battery-powered equipment compared to fuel-burning equipment.
Several categories will have exponential growth of battery-powered units sold along with exponential decline of fuel-powered units:
- vehicles,
- small engine equipment, and
- electricity generators at all scales.
Virtually all sales projections for vehicles fail to anticipate the cost reductions that sodium ion, lithium air, and future battery advances will bring. Instead, most published models assume nickel-based lithium-ion batteries (or ones with similar cost and performance profiles) will continue to be the dominant technology through 2035.
Look Forward, Not Backward
Lead acid batteries were introduced in 1859 and as of 2023 were still used in the vast majority of vehicles sold. The high cost of chemistries with better performance, such as lithium-ion batteries, has limited their use to high-value applications. The iPhone, for example, was made possible by the lithium ion battery technology commercialized in the early 2000s. Before that level of battery performance was available, devices like the iPhone could only operate for a few minutes.
Will better batteries become more affordable and more available, and if so, what are the implications?
I predict that the unit economics and performance batteries will greatly improve in the next decade because one or more of the following will happen by 2030:
- Sodium ion batteries (with positive electrodes containing sodium, iron, carbon and nitrogen) will be proven in vehicle designs and will be manufactured in volume in China.
- Factories to produce sodium ion batteries will start to be built in the United States.
- A public company will commercialize a practical and cost-effective air battery for electric vehicles, including air taxis, that require light-weight batteries.
- A very low-cost battery chemistry (cheaper than the sodium ion chemistries announced in 2022 and 2023) based on sodium, zinc, or iron will become the standard for stationary storage (batteries that can be heavy and large).
- A solid electrolyte material will be proven in a C sample battery for a major automobile manufacturer.
My battery predictions have two important implications.
Implication #1: Changes in Company Revenues
Companies that sell equipment that burns fuel will see revenues decline sharply. Sales of internal combustion engine vehicles and two-stroke engine equipment will collapse. Companies that provide materials for the winning battery chemistries and that produce battery components and manufacturing equipment will increase revenue.
Implication #2: Abandonment of the Public AC Power Grid
Solar power is abundant, affordable and reliable, but diffuse and intermittent. Without batteries, you can’t count on solar electricity being available when you need it. But with good batteries, you can use solar electricity any time. If the relative cost of solar electricity stored in battery systems becomes lower than the cost of electricity generated from coal, uranium or methane, the economic value of the public power grid will diminish.
Due to net metering policies in some jurisdictions, electricity generated by solar power is already cost competitive with grid power, even before factoring in solar tax incentives. Net metering allows excess solar energy to be sent to the grid in exchange for a credit on future electricity bills. From a financial perspective, it’s as if you have a perfect free battery in which you can store electricity and use it in the future without any energy loss.
The only reason solar power can’t compete directly with grid power is that batteries are too expensive. But if battery costs decline by one order of magnitude, solar power with battery storage will outcompete grid power — even if utilities are able to get rid of net metering.
Grid electricity is available for approximately $0.20 per kWh as of 2023 in Maine. The average residential customer in Maine uses about twelve megawatt hours of electricity per year, or about 33 kWh per day. So buying grid electricity costs about $2,400 per year for an average Maine residence. (The average residential customer in Maine does not heat with electricity.)
Because sunlight arrives only during the day, and mostly during the summer, most Maine households can’t rely on it. They need the grid to supply electricity at night and during winter. But if you add about 330 kWh of battery storage (about ten days of usage) so that you can count on having power whenever you want, then you can replace grid power with solar.
Generating 12 MWh of electricity per year (mostly in the summer) requires a 9 or 10 kW solar array installed in southern Maine. Increasing power production and energy storage to provide a margin of safety, an 11 kW solar array and a 350 kWh battery would be sufficient to provide all the electricity a conventional (i.e. not heated with electricity) home in Maine will ever use with no grid power required. For homes that heat with electricity, more solar or more battery capacity would be required to account for the higher electricity demand in the winter, when solar power is less available.
An 11 kW solar array would require about 800 square feet of sunny exposure. The size and weight of a 350 kWh battery bank depends on the volumetric and gravimetric density of the cells. A 350 kWh LiFePO4 battery system would take up approximately as much space as a 275 gallon heating oil tank and weigh a little bit more. (But a full heating oil tank has a lot more energy storage capacity than a battery: 275 gallons of heating oil is about 11,275 kWh of energy, almost as much energy as the average home uses in electricity each year.) Unlike a heating oil tank, a battery receives an energy delivery every day the sun shines.
Once you can lease a 11 kW solar array and a 350 kWh battery for $2,400 or less per year, buying grid electricity will no longer make economic sense for many residential electricity customers in Maine. Currently, you can buy an 11 kW solar array for approximately $2.95/kW or $32,425. Residential battery energy storage systems cost about $700 per kWh installed, so a 350 kWh battery bank would cost about $245,000. At that price, you would expect to pay about $24,500 per year for a solar with battery system lease at an interest rate of 8%. Battery costs must come down by an order of magnitude before solar with batteries will disrupt grid power.
Public policy currently allows Mainers to use the grid as a free battery. This drops the cost of solar electricity to less than the cost of grid power in Maine, depending on interest rates and solar installation costs.
The above chart shows that the cost per kilowatt hour of solar electricity you generate yourself varies from slightly more than 20 cents to less than 10 cents depending on the installation cost and interest rate. (Note that this doesn’t include the value of tax credits, which make solar power much cheaper than grid power.) This chart assumes that you install an array in southern Maine, take out a 30-year mortgage to pay for your solar array, operate it for 30 years, and participate in a net energy billing program that gives you full value for every kWh you generate. The kWh rate is determined by taking the annual payments for a 30-year solar loan and dividing that payment by the number of kWh your array produces per year.
A solar array in Maine likely produces much more electricity than a house needs during summer days, much less than a house needs during winter days, and none at night. Without net metering or batteries, you can’t derive full value from all the solar power you produce that doesn’t match up exactly with the amount and timing of your own demand for electricity. But if you connect your array to the public power grid, you can send your excess solar power to your neighbors.
Net metering allows you to bank solar electricity credits, so you can derive value from every kilowatt hour you generate and share with your neighbors. Under this policy, kilowatt hours you send to the grid entitle you to credits for the kilowatt hours you take from the grid. Full net metering means sending 1 kWh gives you a full 1 kWh credit. Maine has a full net metering policy, but some states have net metering policies that are less than full: you might only get 0.5 kWh credit for every kWh you send to the grid.
Many jurisdictions are deciding that full net metering policies are too generous because with tax incentives the price of self-generated solar electricity is far below grid electricity. That leaves ratepayers who don’t install solar paying more for their power. California, for example, reduced the value of its net metering program by 75% starting April 15, 2023.
Reducing credits for net metering slows the adoption rate of solar but increases the adoption rate of battery storage. Once the cost of a real battery falls below the cost to use the grid as a virtual battery, there will be an economic incentive to stop buying power from the grid altogether. Solar stragglers will be saddled with stranded costs to pay for unnecessary electrical power grid infrastructure.
Few observers are predicting that battery prices will fall by an order of magnitude by 2035. But if they do — that is if battery storage costs can come down from $700 per kWh to $50 per kWh installed — that would reduce the cost of a 350 kWh battery from $245,000 to $17,500. If we assume a lithium air battery can be successfully developed to store 1.2 kWh per kg, a 350 kWh battery would weigh 643 pounds, about one quarter as much as a full heating oil tank. As I write this, lithium costs about $11.35 per pound and iron is about $0.05 per pound, so raw material costs for a battery that size might be somewhere between $7,300 and $35.
We have seen more extreme cost reductions in hard drives ($9.2 billion/TB to $14.30/TB) and solar panels (from $105.70/W to $0.20/W). Achieving similar results in battery cells is possible by making the best use of the iron, sodium, zinc, chlorine, phosphorus, carbon, oxygen, nitrogen, hydrogen and other plentiful elements that Earth provides. Good batteries will transform the energy market and strand investments in public electricity grids as people stop buying grid power.
The public grid in the United States is a 60 Hertz three phase alternating current (AC) continent-spanning machine. Cheaper solar power will disrupt the power grid, because AC power is so inefficient in a world where solar provides most energy. Solar photovoltaic panels, batteries, LED lighting, computer chips and memory, and ECM motors are all native direct current (DC) devices. Plugging DC devices into an AC grid is pounding a square peg into a round hole.
Here’s what happens to power in many homes that have solar and battery back up today:
- A solar panel provide DC power.
- A micro-inverter converts most of that power to AC and loses some to heat and sends AC power to the battery.
- A rectifier in the battery energy storage system rectifies most of that power from AC to DC, loses some to heat, and uses the remaining power to send direct current into battery cells to drive electrolytic chemical reactions to store energy.
- When power is needed, the battery sends DC power to an inverter.
- An inverter converts most of that power to AC, loses some to heat, and sends it to loads.
- A rectifier on devices like LED lighting rectifies power from AC to DC, loses some to heat, then uses DC power to produce light.
Using a DC distribution panel instead of an AC distribution panel, this can all be simplified to:
- Solar panels provide DC power to the battery.
- The battery provides DC power to LED lights.
- LED lights illuminate.
Once solar battery power reaches cost parity with grid power, then distributed power generation using solar panels makes a lot more sense. That means cheaper LED lights, computers, and motors because they don’t need to include circuitry to deal with AC power. A 48 V DC standard will allow most devices to run directly from a battery. Micro-grids consisting of a few solar panels and a few batteries will be safe, affordable and convenient. Homeowners and small business owners will advocate for the right to install small solar micro-grids themselves without hiring electricians or pulling permits. A mesh of solar micro-grids will be easy to interconnect to provide higher reliability than AC grid power.
Block-chain technology will allow smart battery energy storage systems connected to a mesh of DC micro-grids to buy and sell power. Rather than paying $0.20/kWh for grid power, one of your smart networked batteries could bid $0.19/kWh to pay for power to get charged up and one of your neighbor’s smart networked batteries could sell that power to you. A third-party escrow agent would participate in the transaction to verify that power was exchanged before releasing payment.
A centralized AC grid must synchronize the signal alternating its current. A mesh grid does not require signal synchronization. In a DC mesh grid, each battery energy storage system can be directly connected to many other battery storage systems. By making an electrical connection and varying the voltage on each end of the connection, each battery can send or receive power. Switches can open or close to send power in many different paths through the mesh network. Not using AC simplifies the network design, since phases don’t have to synchronized to send power.
Smart batteries on a DC mesh grid can monitor voltage and current to determine which of them are sending power and which are taking power. AI algorithms built into each smart battery energy storage system can use a standardized block chain to record the physical state of the mesh grid and maintain a shared financial ledger in a way that is distributed, verifiable and tamper resistant so that a peer-to-peer energy market can be used to allocate resources according the needs and desires of battery owners.
Micro-grid power block chains can scale from inside a device (where various components are bidding for power and a control algorithm decides how to allocate power among components), to inside a home (where appliances are connected to a common distribution panel and bidding for power and a control algorithm decide which appliances get power), to inside a neighborhood or campus (where battery energy storage systems are bidding for power in a mesh grid), to towns, states, regions and larger networks. The analogy is how information systems based on the transmission control protocol / Internet protocol can scale from inside one computer to the entire Internet spanning the globe, which resulted in profound changes to our telecommunications industry.
Almost everywhere on Earth, it will become more expensive and more risky to use AC grid power than to use a battery charged with solar energy connected to a mesh grid controlled by AI algorithms using block chains to allocate scarce energy and power resources.
When Will the Battery Revolution Begin
If any of the thousands of pairs or redox reactions now being investigated can reduce the cost of batteries by one order of magnitude by 2035, we can expect massive changes in how energy is generated and used.
Until 2020, the battery market was relatively small and the general consensus among manufacturers was that existing chemistries were good enough, as evidenced by the fact that most vehicles still used lead acid batteries. Annual capital investments in battery technology were relatively modest until 2020. Beginning in 2021, more investors began allocating more capital to improve battery technology, including the US Department of Energy. We should begin to see the impact of some of those investments starting in 2025 when factories begin making batteries at scale.
AI and block chain technologies are still in their infancy, despite persistent predictions since the 1950s that we would make faster progress. The recent resurgence in interest in computer science as an undergraduate degree gives some hope that progress in these fields might accelerate so that a market of mesh-connected micro-grids could start to replace the centralized power grid by 2035. “Smartless” stand-alone batteries will have much less value than smart networked batteries once we have software that can create a resilient and robust peer-to-peer electricity market.
One indication that we are entering the explosive growth phase of the battery revolution will be a global fire sale on lead acid batteries. Another indication will be an announcement that all major automobile manufacturers are moving away from 12 V to 48 V for low-voltage power in vehicles, accompanied by a fire sale on 12 V accessories. A third indication will be panic among utility executives and commissioners as they realize that the demand for grid power is evaporating much faster than anticipated.
By 2035 we will probably have at least one order of magnitude more batteries being produced in the world, and very likely those batteries will be one order of magnitude more affordable on a cost per kWh basis compared to batteries being sold in 2023. That, combined with advances in AI, will unlock many opportunities to create wealth for investors who make wise choices in 2023, 2024 and 2025.