Renewable energy battery backup

The various renewables promoters are suggesting that the UK should aim for '100% renewable energy by 2050' on the strength of claims that new battery storage technology will soon be available to overcome the intermittency of wind and solar energy. Let's look at the feasibility of this idea.


  • Intermittency is the major drawback of renewable energy
  • Outages can last several weeks
  • A week's backup with the best existing batteries would cost more than 30 moon missions
  • In spite of this, promoters claim that economic energy storage is under development
  • The onus must be on renewables operators to provide the promised storage
  • Contracts must not leave the consumer to pay for any failure to do so 

Why renewable energy storage would be needed

Wind output can be low for a week or more.Solar PV provides a burst of energy centered around mid-day. This may extend for as long as twelve hours in summer, but the maximum output is only achieved for a few hours. Cloud cover can reduce the output by a large factor, hence energy yields differ from day to day. In midwinter, output is typically one tenth of that in summer.

Hence we see here that firstly, backup must cover the period between sunrise and sunset every day. That is the minimum backup requirement. The worst case requirement is that of winter, when two or more months may suffer near-zero solar power.

Wind also suffers intermittency, although it follows a rather different pattern from solar. Greatest outputs tend to be in the winter and spring, with less in summer and autumn. At any time of year, a high pressure region settling over the UK can result in several days with little or no wind. In some instances this has extended to three weeks of calm.

The Grid has little inherent storage capacity. For the most part, generating capacity has to be adjusted in real time such that supply equals demand.

Renewables can usefully serve as part of Grid capacity, up to a point. Provided that the percentage of renewables is significantly less than that of other power sources in the mix, the other sources can quite simply be adjusted to make-up for the changes in renewables output. This arrangement is what we presently have, and it works well.

However, when the requirement becomes one of entirely replacing other energy sources with renewables, that option no longer exists. In that case, there HAS to be some way of storing the renewable energy for use when the wind doesn't blow. 


German electricity  production

UK electricity production

UK historical electricity production

How much power?

UK electrical requirements are for a base night-time demand of about 25GW*, rising to 35GW on a summer day, and up to 55GW on a cold winter day. Of that, about 8GW is available from nuclear, plus a couple of GW from hydro and international feeds.

Thus, if we are to eliminate fossil fuels that leaves an amount ranging from 15GW to 45GW to be supplied from renewables. Or, from batteries when there is no wind or sun.

* To give an idea of scale, a gigawatt is essentially a million one-bar electric fires, or ten million traditional light bulbs.

For how long?

The backup requirements for both wind and solar energy types require that for complete security of supply, other sources, or stored renewable energy, must be able to stand-in for several weeks at a continuous stretch.

Logistically, if we wanted to be genuinely 100% renewable, we'd need battery backup to cover about a month of cloudy skies and nil wind. However, such extended outages only occur once in a while, so to be more realistic let's specify a week's battery backup, and agree to use fossil fuels to cover longer outages. After all, that shouldn't happen too often.

So as a ballpark, a week's backup is what we'll settle for. The UK baseload, minus nuclear and hydro, we could take as 25GW on average. So, we need a battery with a capacity of 25 x 24 x 7 gigawatt hours. 4,200GWh.

How large would such a battery be, and how much would it cost?

Traditional lead acid

A car battery stores about one millionth of a gigawatt-hour,and costs maybe £50 wholesale. It is probably the cheapest option. Lead acid technology is however not ideal for this kind of duty owing to its problems with deep discharge and limited cycle life. Lithium would be a better option, although more costly.

The lithium based Tesla Powerwall. Lithium

Turning towards batteries actually intended for this kind of purpose, the lithium-based Tesla Powerwall stores 14 kWh, or 14 millionths of a gigawatt-hour, and costs $5,500 USD at one-off prices.

The individual cells typically used to build such packs are known as the 18650 style, and cost from £1 up in bulk quantities, although the quality brands will typically be £2 to £3 each. They each hold around 2.5Ah at 3.7 volts, which equates to around 10 watt-hours. Thus a Powerwall-equivalent pack would contain 1,400 such cells, at a cost of £2800 if the unit cost is £2. This fits in quite well with Tesla's retail price for the Powerwall, so let's just assume that we couldn't undercut Tesla's prices by all that much even if we built our own.

To give it its due, the Powerwall is actually quite a cost-effective product for home electricity backup. However, Grid-scale backup involves entirely different amounts of energy storage. It's the sheer scale of such a project that's the issue.

There is also a larger Tesla product, the Powerpack, which according to Wikipedia  has a per-kWh cost of around $400.

Flow batteries

Another possibility being widely touted by the various Green Tech websites is the Flow Battery. Using liquid electrolyte stored in external tanks, this does lend itself to larger deployments. Figures for actual cost of these units are quite hard to come by, but the lowest projected cost I've found for a unit in-development is $250 per kWh, or about $3000 for the equivalent to a Powerwall in storage capacity. That might come down with development of this relatively new technology. For the moment though, all known available flow batteries are more costly per unit storage than lithium.

OK, Tesla it is.  

So, let us assume that we could talk Elon Musk down to $4000 each for a bulk purchase of Powerwalls. How many would we need for our week's Grid backup?

14kWh is 14 millionths of a GWh, so to supply 4,200GWh for our week of no wind or sun, we would need 4200/14 million of them. That's 300 million Powerwalls for a week's Grid backup.

Thus, every person alive in the UK has to buy five Powerwalls and we're done. No problem. Of course that includes OAPs, the unemployed, and children on pocket money.

In national budget terms, the cost would be 300 million times $4000, or $1.26 trillion. For the sake of argument, call it a round trillion pounds Sterling.

Bear in mind that's for a week's backup. If we wanted to go the 'really, really green' route and ensure that we never (well, hardly ever) had to resort to burning stuff dug out of the ground, the cost would be more like four trillion.

How much is this amount of money worth, in relative terms?

At current costs, instead of a week's Grid backup, we could put sixty British astronauts on the Moon.When we talk about such large amounts of money it's often difficult to visualise the kind of sum involved. Therefore let's compare it with a very large, very costly endeavour we all know of: The Apollo program.

According to Wikipedia, the entire Apollo project cost $24 billion US Dollars, 1973 value. The equivalent in today's money would be somewhat over $170 billion dollars.  We could call it £200 billion for the sake of argument. So, for the cost of a single week's Grid backup using batteries, we could finance at least five Apollo programs. The original Apollo program put six crews of two astronauts each onto the surface of the moon. There were also three missions which visited the moon but did not land, but for the sake of argument let's call it six.  So, six missions for £200 billion.

Think about this carefully for a minute: The cost of just one week's worth of Grid backup, using technologies known to exist today, is equal to that of thirty moon landings.

In reality the cost of 30 missions would probably be less than five times the cost of six, because your development costs would only arise once, and would be be spread over a larger number of launches. It gives us a good enough comparison though.

If we decided that we needed four weeks' worth of battery backup, then the cost would equate to at least 120 missions to the moon.

Of course in reality, we might decide to spend the money on something other than moon landings. Your call on that one. I'm just trying to illustrate the scale of the money involved, though.

If the costs seem astronomic, why is battery storage being promoted?

One possibility is that the battery backup idea may be a gambit, played by renewables promoters in order to secure increased government funding for their products. By talking as if the development of battery backup is almost a done deal,  politicians may be convinced to offer favourable contract terms, untroubled by concerns over intermittency. 

Intermittency is after all the single greatest drawback of renewables, and IF it could be solved, then renewable energy would be a far more valuable commodity on the world energy markets. It might even be seen to entirely replace fossil fuels in some roles. Thus, to claim that energy storage will soon be available, and at a viable cost, is nothing short of playing an ace of spades in the energy game. Is that ace genuine though, or did it come from the gambler's sleeve? That is the question.

Existing supply contracts require that all energy generated by renewables has to be paid for, whether is used or not. These arrangements were provided since historically it was recognised that the lack of energy storage could otherwise leave fledgeling renewables operators unable to break even. However, with renewables promoters now stating categorically that energy storage will be available, and soon, the justification for providing such contracts no longer exists.  

Indeed, if renewables operators were able to negotiate further such contracts, the failure of the promised storage to materialise would hurt the consumer, but be largely unimportant to the equipment operator, for whom it would be business as usual generating electricity that might or might not be wanted, and getting paid anyway.

That would be a dangerous situation to get ourselves into, since a poorly-worded contract might land the taxpayer with still having to pay for unused energy from wind turbines, solar farms etc for which no energy storage exists or will ever exist. Even though the equipment was bought on the strength of energy storage being promised, it might prove hard for governments  to back-out of such contracts. 

Ways to handle this situation

The important aspect of this, therefore, is not to let our governments get us into open-ended contracts for renewable energy infrastructure, as has so often happened in the past. Instead, either make it a condition of payment for such hardware that energy storage shall be developed, or else only offer to pay a strike price for dispatchable renewable energy.

That is, pay only a guaranteed minimum price for energy which meets an identified demand. Thus, if the operators have energy storage as promised they will be able to sell their energy at a profit as and when it is needed. In which case everyone is happy. If it transpires that they do not have and cannot develop storage though, then they will have to take whatever price they can get for their energy at the time the it becomes available. Which, if all demand is already being satisfied, might not be much.

With such an agreement, if cost-effective storage is developed then it's win-win for everyone. Failure to develop cost-effective energy storage would only hurt the operators who promised us that storage in the first place. Instead of hurting the public, the Grid operators, or other energy providers, as happens with unconditionally guaranteed prices.

There are indications that the Westminster government has already set it sights on the second approach, of offering strike prices for dispatchable energy, thereby making intermittency a problem for the equipment operator to deal with, rather than for the user. 

I've not yet seen any indications that Holyrood is following suit, but let's hope they see the sense in such an arrangement. 

Images courtesy of Wikimedia Commons and Euan Mearns