Over the last two weeks, we wrote a lot about the “Kombikraftwerk2”, 100% renewable power scenario. While the study focused on the technical feasibility, one of our articles questioned the economics of the described storage requirements. But is storage really that expensive?
Last week, my colleague Craig Morris wrote about the affordability of the storage needs described in the Fraunhofer IWES study (see above). While he considered the use of power-to-gas (P2G) in the scenario reasonable, he had this to say about the role of batteries on the grid:”55 GW [of batteries] running below 1% capacity. What a waste! We need no cost estimate to see that this item is off kilter”.
IWES 100 percent
I was not willing to accept this statement without crunching some numbers by myself. As it turns out, we actually do need a cost estimate for battery storage and the result is not that terrible in my opinion.
The role of batteries in the scenario
Today we usually discuss the cost of batteries in the context of energy storage. How much does it cost to store a unit of energy (kWh) and how much of that energy can we retrieve whenever it’s needed? The answer to this question mainly depends on the batteries investment cost and lifetime (in years and cycles).
The low capacity factor of the entire battery park ( lt;1%) in the IWES scenario is an indicator of a rather different role for batteries. A look at the wonderful animated presentation of the scenarios load curve reveals that batteries are mainly used to provide power whenever there is a regional gap between dispatch-able power generators covering the residual load and the output from variable renewable sources. Rather than storing high volumes of energy, the batteries in the scenario act as fast reacting distributed power buffers.
The right batteries for the job
There are several battery technologies out there that are suited for the job in question. All of them come with their unique set of pros and cons; the downside of the most intriguing ones is usually that they are still in an early stage of commercialization, or even in the lab. There is however one type of battery that appears to be well suited for the job and has been produced on a multi-GW scale for decades: the good old lead acid batteries, also known as the “car battery”.
It’s no secret that car batteries are quite cheap. They usually cost only about 100€/kWh and even the more expensive high performance ones are in the range of 140-180€ / kWh. Their downside is however that they are relatively heavy (that’s the lead part) and that they have a relatively low cycle (200-800 cycles) and calendar lifetime (3-12 years). Both downsides are however not a major issue for the stationary application we are looking at.
Batteries for balancing power are a minor expense
A battery is, however, not a battery system, so let’s be conservative and assume that a system would cost about 300€/kWh (this paper puts it at 50-300 USD / kWh). This puts the total cost for building the entire 55 GW of battery storage at 16.5 bn. euros.bn. € / GW GW Lifetime bn. € Invest bn. € / a " battery fee" in cent/kWh Lead acid battery 0.2-0.5 55 (4) 8 - 12 11 - 27.5 0.9 - 3.4 (6.9) 0.17 - 0.64 (1.3)
Now, if we assume an average lifetime of just 8 years, the battery park would cost about 2 billion euros per year, which when evenly spread across the entire electricity consumption of 524 TWh, amounts to about 0.4 cent / kWh.
Granted these are very rough and simplistic cost estimates. No financing costs are included and despite making rather conservative assumptions for both investment cost and lifetime, I am properly off somewhere. I am however fairly confident that the cost of batteries for balancing power for the customer would be below 1 cent /kWh. While this is a factor, I would not say that it's enough to make a 100% RE power supply unaffordable.
This price tag for getting rid of large centralized power stations must however be considered in the context of grid fees and not simply as an additional expense. The cost of having batteries on the grid would to some extent be offset by a reduction in cost for conventional balancing power. Considering that current regional price differences for residential grid fees (in German, see the map on the last page) of more than 3 cent / kWh are barely ever talk about, I doubt that this make-it-work requirement of a 100% RE power supply would be a deal breaker for many. (Thomas Gerke)
Edited on September 9th - Table with the data on the calculation added.