By Kevin Kilty

Figure 1. An energy storage scheme with complexities galore. High temperatures, exotic materials, and extreme machine demands. (h/t Chaamjamal)

Air compressed and shoved into old mines and caverns; mass hauled uphill, and let down again; heat stored in mass. New, advanced energy storage schemes? Well, no, these are from the last big energy crisis 1968 to 1982.[1]  However, anyone paying a modicum of attention to either the professional or popular literature must have noticed lately an inordinate number of energy storage schemes being proposed that run the gamut from simple to bizarrely complex (see Figure 1).[2]  I see this as an implicit admission that renewable energy systems cannot work on their own without an array of ancillary services supplied by someone else.

Recently an acquaintance who works for a major U.S. utility sent me brochures describing several schemes, and asked what I thought of them. Two store energy in mechanical ways, and one is a thermal storage system resembling the one profiled on WUWT a few weeks ago. As the time scale to switch to 100% renewables is constantly being advanced on us, even as we only begin to recognize the unsuitability of renewables alone, my acquaintance said, “We are going to waste a lot of money in the next ten years.”

Energy storage

Traditionally we stored energy for our prime movers, lighting, and home heating as agricultural products — hay and grain, candles and oils, and wood respectively. The industrial revolution with its greater scale of energy use continued this same scheme. Instead of feeding old Dobbin (see Figure 2) his hay and grain and putting him into harness, though, we draw upon the stored feed for the thermal plant and have it follow demand for work during the day. If a person only looks backward on the grid as far as the thermal plant, the current electrical grid looks a bit like a just-in-time system, where energy is converted in careful time balance with demand. It is undeniable however, that energy to convert is drawn from stored fuels. Just as grains are a superior source of stored energy for the horse, fuels for combustion are a superior source of stored energy. At large scale energy is surprisingly difficult to store effectively by other means.

Figure 2. Old Dobbin, the prime mover. Give him oats, hay and water. He’ll do 550 foot-pounds of work per second all day. Cecil Aldin circa 1915.

There are four fundamental problems with energy storage schemes; the energy scale, size/material scale, work scale (or how the universe works), and operations. Let’s examine each in the simplest manner possible.

Energy Scale

Nowhere in this section will I worry about how the stored energy can be turned into work in practice — I am just comparing raw calculations of energy. Take a kilogram of diesel fuel and equate its lower heating value to other forms of stored energy.

The lower heating value of one kilogram of diesel fuel is roughly 43,000 KJ. If we try to store the equivalent energy by raising mass, which is what many energy storage schemes attempt, it is the equivalent of one metric ton of mass raised to a height of 4,300 meters. Let’s now compare it to stored rotational energy. One-hundred forty kilograms of diesel, roughly what goes in the tank of an over-the-road semi-tractor, equates to the rotating energy of the turbines and generators of a 3000 MW nuclear plant in operation. Store it in an electrical field? This 43,000kJ is the energy in the electrical field of three and a half million 5 volt, 1 farad super capacitors. Thermal energy? The equivalent is one ton of air or rock at 43C (77F) temperature increase. What comes closest is chemical energy of rechargeable batteries. Our 43,000 kJ  is approximately what energy thirteen good quality automotive batteries or 1,300 3.7 volt lithium cells store. The 50kWhr Tesla battery, composed of Panasonic 4680 cells, is about 4 kg of diesel.[3]

Combustion releases a lot of energy which raised mass, spinning mass, electrical and magnetic fields, or heated mass cannot match. Only chemical energy comes close.

Size/Material Scale

A convenient way to illustrate this is by using the electrical energy destroyed daily by 1000 Americans, or what is equivalent, the daily electrical energy used in about 300 American single family residences. This amounts to 9,000 kWhr of electrical energy[4], which is the equivalent of raising three unit trains with attached locomotives, of 10,000 metric tons each to a height of 100 meters. It is the turbines and generators in six nuclear plants spinning at nominal output.

One of the mechanical schemes I was sent uses 35 metric ton bricks, six at a time and raised nominally 50 meters, and the other uses 357 metric ton railcars loaded with rock moved by cables up and down a steep incline. These systems sized to 9,000kWhr amount to 2,000 bricks, or about 70 railcars, respectively, and either one requires around 2 hectares of land area.[5] While one of the modular systems shown in Figure 3 may be a novelty in some towns, dozens might be considered blight.

Figure 3. The Energy Vault, a European storage system uses 35 metric ton bricks stacked to store energy and unstacked to release it again.

As chemical storage systems seemed to compare most favorably in considerations of energy, we should see how they compare here. To store 9,000 kWhr we need something like 270,000 Panasonic 4680 cells, with a total mass of 35 metric tons; or, 225 metric tons of lead-acid batteries.   

The thermal storage system is the worst of all. From its description of separate hot and cold storage sides, with mass ratios of it requires at least 240 tons of rock, all suitably encased in steel with insulation.

The way the universe works

When I refer to “the way the universe works” I mean obeying the laws of thermodynamics. The systems which suffer most acutely from limitations imposed by thermodynamics are thermal and chemical storage systems. Mechanical systems, aside from friction, represent 100% available energy — that is, all the energy they contain is available for work.

The problems with thermal and chemical systems arise from what we call “irreversibilities” — things that are impossible to undo without expending more work in the process. There are two categories to consider: internal irreversibilities, which are those within the power cycle of the system; and external irreversibilities, which are those arising from the interface between our storage system and that machinery which it serves.

The inefficiency of the compressor and turbine contribute most to internal irreversibilities of the thermal power system. I used an air-refrigeration cycle (see Figure 4 for the ideal) to analyze this proposal with assumed 90% efficiency for both turbines and compressors to find a round trip efficiency of about 68%. At 80% efficiency for both compressor and turbine it drops to about 46%. But this applies only to a short period at turn around of recharge to discharge. As time drags on the efficiency goes down. Thermal systems are notoriously ill-suited to storage and then to recycle energy because of the way the universe works.

Figure 4. An idealized Joule cycle using air as its working fluid. COP is the coefficient of performance of the charging cycle, while epsilon is the efficiency of the reversed cycle providing power. Note that 1.75 times 57% is 1.0, indicating the ideal round trip cycle is 100% efficient. This is nowhere near the cycle efficiency as operated with real equipment on a real schedule.

What contributes most to internal irreversibility in batteries is that they depend on molecular diffusion through an electrolyte. The force that drives diffusion is an internal electrical potential at the electrodes known as an overpotential. It raises the voltage needed to charge batteries, and decreases the voltage available externally upon discharge. It is absolutely necessary to drive diffusion but at the cost of turning available energy into heat rather than work. The overpotential grows with current so that when large amounts of current are called upon to recharge a battery or turn a motor quickly, the overpotential and work wasted both increase too. Also, as temperatures in the battery rise, unwanted side reactions occur. These perpetually degrade the operation of rechargeable chemical storage systems.

The principle issues behind external irreversibilities are the efficiencies of the various ancillary equipment involved — electric motors and generators, and especially heat exchangers, all of which are required to make an energy storage system operate as a useful utility. Advanced batteries also require system monitoring electronics, buck/boost power supplies, control systems both for charging and discharging, and heat exchangers for cooling as well.

Operational Issues

Operational issues means problems like the following: making a storage system serve many purposes which are at odds with one another, or operating a system in a mode where it has to make constant reversals of cycle, or operating at a slew rate it can barely meet.  For example, a need to store energy for a long term power outage which suggests always having a full reserve is incompatible with the need to always have capacity to absorb or supply energy in order to provide ancillary services for grid stability. Mechanical and thermal systems are incapable of quickly supplying energy for sudden grid disturbances and need a grace period of seconds perhaps to respond.[6] Thus using them in a grid bereft of inertia, as we are trending, will require other as yet unspecified systems to supply this inertia.

Cycling power systems quickly can end up stranding stored energy in a way that it has no path except to the dead state. No matter how one may try to limit this problem, heat is going to leak away from any thermal storage system and chemical reactions will occur within the electrolyte of a battery to deplete its charge even as it sits idle. Repeated starting and stopping of thermal systems leaves heat energy stranded in the heat exchangers, turbines, compressors, and this heat ends up in the dead state unavailable for useful work. Think for a moment about turning on a shower at home and waiting for it to heat up sufficiently to be comfortable. We waste warm water while we wait. Then there is shutting down the shower when all the piping is at temperature and leaving it to cool back to home temperature. All this leads to heat being dumped into the dead state that is the house environment.[7] Finally, consider that the thermal system is not attached to reservoirs that can maintain constant temperature, but to finite stores of heat which decline in temperature the moment we begin to use them. Likewise batteries are a finite store of charge where their voltage declines as soon as we begin to demand current.

Then there are practical matters, like how does the stacking of 35 ton bricks work (Figure 3) when covered in ice and snow?


By the time we consider all the factors I have mentioned most of these storage systems will have to be overbuilt by factors ranging from one and a quarter (1.25) to four (4) or more. Even then they may not supply a full range of ancillary services the grid requires. I did not address the problem of cost, but every problem I did address will eventually impact costs. There are many reasons to believe my acquaintance is correct when he spoke of wasting a lot of money. It took money spent over a century to learn the systems engineering currently built into the grid. It will take a lot of money to duplicate all this for a completely re-imagined grid in a decade. As Petr Beckmann, who was a stout proponent of nuclear energy, said forty years ago, “soft technology will not be America’s energy salvation.”

One more point seems pertinent. A decade ago we were all lectured that the environment would be best served as we de-materialized society. That is, we would reduce our environmental footprint as we used less matter; reducing the energy inputs and associated pollution that processing materials requires. Renewable energy with the low energy density of its sources, and the storage it requires and using many advanced and rare materials, are taking us the other direction. Instead of depending on dispatchable energy sources which can supply a kilowatt per kilogram of material, we are proposing systems which provide a mere watt using tens of kilograms. We will re-materialize society. 


1-As just one example, in the early 1970s one alternative energy wind machine used a “race track” design wherein railcars with airfoils would traverse an elongated track oriented at right angles to the prevailing wind direction. My last senior design class, one group studied and made a preliminary design for a wind energy system that, again, used a “race track” concept, but with even more problematic engineering issues than the original. Compressing air in caverns and mines is straight from the 1970s.  

2-Chaamjamal, a regular commenter here, pointed me to this scheme which contains an element of about every technical problem imaginable.

3-I can find little about the specifications that really matter on this battery. People seem fixated on its dimensions, 46x80mm, rather than C-rating, useful energy stored, and slew rate. Allegedly it involves 960 3.7V cells each with 9Ahr C4 rating. This doesn’t work out to 50kWhr, and the blog-o-sphere is full of impossible recharge rates and other mythology.

4-9,000 kWhr = 300 homes at 30kWhr each per day. The average home uses more energy in natural gas to provide heating and domestic hot water. Here the rate is two Therms per day at 29 kWhr per Therm. Twice the amount of electrical energy used.

5- The GravityLine system which uses rail cars filled with rock is much like pumped hydro storage but solves the problem of not having spare water everywhere. However, it still suffers the problem of there not being suitable terrain in many places.

6-The slew rate and inertia in power systems has to be matched against the time response of the grid to its disturbances in order to keep blackouts at bay. The margin against ungraceful behavior is often measured as the ratio of the amount of rotating kinetic energy (turbines and generators) against the rate they nominally deliver power. So, for example, a nuclear thermal plant, nominally producing 2,500 MW may have rotating energy of 6GW-seconds, making the ratio of the two about 2.5 seconds. This provides adequate time margin for automatic control to maintain frequency and voltage when a generator goes offline, or a sudden load appears.

7-The dead state is an engineering concept. It is a state at ambient temperature, atmospheric pressure, and zero electrical potential where energy finally arrives too depleted to do further work. You have heard of the heat death of the universe? The dead state is the funeral parlor.

via Watts Up With That?

2021 May 29