By Mike O’Ceirin
Much has been written about using energy storage to stabilise renewable energy. This article uses actual data to present a theoretical answer to what may be achieved. The pattern of demand comes from actual Australian data as does the wind output. This has been put onto a website for public access and to be referenced by this article. The conclusion is that in the Australian environment on our eastern grid using wind and Pumped Hydro Energy Storage (PHES), for every TW hour 500 MW of wind and storage of 22 GW hours is needed for stability to match the existing demand pattern.
Requirements to maintain Status Quo
Australia faces five closures due to age of coal stations by 2034.
|Asset name||Year||Capacity MW||Dispatch 2020||State|
|Vales point B||2028||1320||3.6%||NSW|
Since the Australian demand for electricity on the eastern grid in 2020 was 203 TW hours, renewables need to dispatch an extra 55 TW hours per annum to replace them. This cannot be done just on average. The same pattern of electricity demand must be satisfied. To do this theoretically wind and PHES has been chosen.
This estimate shows that to supply that amount of electricity under Australian conditions on the east coast 27 GW of wind generation and 1182 GW hours of energy storage will be needed.
Capital costs for the wind according to Gencost 20 – 21 (page 17), at $2000 per kW hour would be $54.3 billion plus a large cost in transmission lines. For the cost of PHES see appendix storage costs.
Energy storage is nowhere near as determinate as the renewable energy sources. PHES is a bespoke item which has extreme variation in price. Certainly, there will be a long-time frame for the building of PHES and producing sufficient to meet the requirement. According to this estimate the task of producing such a large amount of electricity storage may well be prohibitive. For the Gencost estimate the PHES would cost $62 billion and a Kidston (see appendices) equivalent $456 billion. Note that this cost has been minimised by assuming willingness to run at critical levels. In a practical world this would be unacceptable.
This is only needed for variable renewable energy since energy storage is not needed otherwise. Specifically, it applies to wind and solar production of electricity. This article concentrates on electricity produced from wind. The technology for the storage is confined to PHES. The reason for choosing that for electricity storage is its capacity and cost.
The estimate relies on data (see appendix Data Source) held on the author’s website Spasmodic Energy. This estimate uses the patterns of electricity production and consumption derived from that data.
First the demand must be known. The demand for electricity on the eastern grid in 2020 was 203 TW hours, on average 23 GW hours per hour. For calculations of the energy demand this is not sufficient.
The electricity demand chart (figure 1) shows the pattern. On any day, demand is greatest in the early afternoon and the lowest in the early hours of the morning. This is expected and must be taken into consideration when calculating the energy storage requirement. It is suggested that estimates such as this should be done against a full year since there is a definite yearly pattern. By that means the low and high points of the year for wind will be included.
Using actual data an estimate has been made by stepping through each actual hour.
This is shown graphically on the energy storage page (figure 2). The pattern of the dispatched electricity is shown in the lower chart superimposed over the actual output of wind for the chosen time period. For each step, the assumption is made that the initial percent can be achieved. The steps to solving this are as follows.
- Is there enough electricity from wind to supply the demand for that hour?
- If so, add any excess to storage less 15% (to account for storage losses).
- If not, then draw enough from storage (there is a 15% loss here as well) to satisfy the demand.
- If storage has gone to zero or below then the run has failed. Reduce the demand percent by one and start again at the beginning.
The output is shown in figure 2. This is a run for the whole of 2020 starting with an initial charge of 350 GW hours and an expected demand of 10%. The demand is an expected percentage of the total demand for that year. As above it can be seen (on the right) that 10% failed as did 9% finally settling on 8%. The result is then displayed in the two charts. The top one shows the charge of the storage throughout the year. In the example it is stable until the end of January and then it falls and keeps on falling until early March where it sits at practically at zero for many days. The storage gradually builds up to become fully charged at the beginning of May.
Please note that the displayed example is static while the actual website is not. There the whole picture can be seen throughout the entire year by scrolling the display. This was found to be quite important since trying to display that amount of information in a static chart does not show what happens. The charge drops again to under 50 megawatt hours in August to again rise to a full charge at the end of August which remains until the end of the year.
The second chart shows what the result is. It is clear that energy storage reduces the amount of usable electricity. In 2020 wind dispatched 19.68 TW hours but it was not stable. If it is stabilised using energy storage the usable wind energy drops to 16.22 TW hours. The cost in this case for stability is a loss of 18%. There is the overhead of the energy storage of 30% and when it is fully charged potential electrical generation is lost. In the energy storage chart figure 2 this can be seen in September (if scrolled).
Why is this so? It seems very extreme. The point is wind droughts are frequent and can be long. These can be found on the website under “wind droughts”. In 2020 many instances can easily be found where the capacity factor drops to under 5%. There is one extreme case where the average was 5.5% for 33 hours. There are others where the drop is to 2%. Despite the fact that the Australian eastern grid is very large, covering many thousands of square kilometres, the wind patterns are larger. The whole of Australia can easily have little wind also too much wind.
To stabilise wind energy generated during 2020 so that it is congruent with the demand 350 GWh of energy storage was needed. That will meet 8% of the demand using the existing wind infrastructure of 2020. The capacity of wind was 8 GW and 8% output is 16.22 TWh. So, for 1 TWh 0.5 GW of generation is needed. As a rule, 22 GW hours of storage are needed to stabilise 500 MW of wind.
As previously stated, five major coal stations will close before 2034 on the Australian eastern grid. Together they currently generate 27% of the electricity on the eastern grid. That is 55 TW hours so applying the rule 1181 GW hours of electricity storage will be needed. At a minimum cost of $52 million per GW hour the total would be $61.5 billion.
Australia has a grid along the whole of the eastern coast which connects generators. That grid is monitored by the Australian Energy Marketing Operator AEMO who records and publishes pricing and performance data at five-minute intervals. That data is used in this estimate. To enable the use of this data on a website it has been transformed to hourly data. Currently there are 150 million data points. Wind is used as the source of the renewable energy which in Australia is spread over a large area.
Pumped Hydro Energy Storage (PHES)
This estimate assumes that this can be provided with sufficient capacity. To estimate the cost the Gencost 20-21 study (page 19) can be used. In Australia there is a proposed large PHES to be built known as Snowy Mountains 2.0 (SM 2). It is projected to be able to deliver 2000 MW for 175 hours. This results in a capacity of 350 GW hours. The default trial figure is the same, but other figures can be applied of course to the storage page of the website.
Gencost gives a cost per kW and that cost is for 48 hours. The cost for 2000 MW on that basis is $5 billion. Matching it to 350 GW hours means it must be multiplied up and the result is $18 million of $2 million per GW hour. The Snowy Mountains 2.0 as previously stated is the same as this in capacity. The cost that has been often stated is $10.4 billion. That makes it far cheaper than the Gencost study would allow, even though the cheapest has been chosen out of that study, since much of the work has already been done. Possibly it is an exception.
Kidston is being built in Queensland in an old gold mine. Compared to SM 2 there are significant differences. It is much smaller and much more costly. The capacity is 2000 MW hours. The cost is $386 million per GWh.
via Watts Up With That?
August 12, 2021