Value chain · CLM-1488
In plain terms
A Royal Society report on storing electricity in Britain, drawing on the Met Office, says a spell of very low winter wind lasting up to a week comes every few years, and a calm summer spell of up to four weeks once every ten years. But in its model of a 2050 grid run on wind and sun, the biggest drain on the store came in the weather of 2009 to 2011, a run of low-wind years. The report warns that studies of single years, or of only the worst spells, can badly underestimate how much storage is needed.
The claim as recorded
The Royal Society's 2023 report on large-scale electricity storage for Great Britain, citing a Met Office review, lists a winter wind drought of up to a week of very low wind speed as happening every few years, a summer wind drought of up to four weeks once every 10 years, and a full day of very low summer wind one or two times a year. It cites a Met Office study that found about a 10% chance per decade of a winter month with lower wind speeds than in the period studied, which in context is the 1980 to 2016 period the report modelled, and it says that prolonged periods of low wind can occasionally last a few years. In its model of a 2050 system run on wind and solar power backed by hydrogen storage, the report finds a very large call on storage in the weather of 2009 to 2011, which it attributes to persistently low wind speeds, and its figure shows the store at its lowest level in that period. It says that a study of the years 1984 to 2006 alone would have found a much smaller store, and that studies based on less than several decades of weather are liable to very seriously underestimate the need for storage, especially those that look only at single years or at selected periods of high stress.
Evidence behind this: T3: a Royal Society policy report with a named report leader (p.92); it is not a peer-reviewed journal article. The frequencies in its Table 1 are the report's summary of a 2019 Met Office literature review, and the 10% figure comes from an unpublished Met Office study. Neither original was read, so both are graded on the report that cites them.
Scope and limits
The frequencies are not the report's own counts. The report says the three event types were identified in a 2019 Met Office literature review; the table gives no separate source for its frequencies, and that review was not read. The 10% figure comes from a preliminary, unpublished Met Office assessment of low winter winds over the North Sea, not Great Britain alone. The report does not define "very low wind speed", list the years in which each event occurred, or give a number for "Every few years". The lengths in Table 1 are upper limits ("up to"). Elsewhere the report says periods of scarce wind and solar supply together "can last up to two weeks" (p.28), which is a different measure from the four-week summer wind drought; it does not reconcile the two. Its point that the largest deficit builds up because "multiple scarce periods can follow each other closely" is shown by a study of Germany, not Great Britain (p.28). The model's demand does not follow the weather: it leaves out "correlations that occur during winter anticyclones when it is cold and wind speeds are low". The report finds this probably safe for the need for storage over decades, but says it "will lead to underestimates of the need for storage on shorter time scales", and relies on the store's contingency to cover it (p.20). So the model understates the storage needed to get through a cold, calm winter spell. The store's energy size and its power are set by different things: the report sizes peak output separately, at 100 GW against a modelled peak residual demand of 88.2 GW (p.26). Figure 13 is one case at 1.3 times demand; the cost minimum is at 1.33 times (p.31). That the 2009 to 2011 low is the lowest point is read from the figure, not stated in the text. Read from the same figure, single seasons also draw on the store heavily: in the first months of the run, which starts in 1980, the level falls from about 93 to about 40 TWh. The report says 37 years is not a full sample of rare weather (p.6), and that 1960 to 1980 had lower wind speeds than 1980 to 2016 (p.19), so a store sized on this record may be too small. Great Britain, 2050, one modelled system; not a forecast.
What it assumes
The model compares simulated hourly wind and solar output for 1980 to 2016 with one hourly demand profile for 2050, 570 TWh a year, based on the weather in 2018 and repeated in every year (p.17, p.19). Supply is 80% wind (70% offshore, 30% onshore) and 20% solar. The Figure 13 case has wind and solar output of 741 TWh a year (1.3 times demand), hydrogen as the only large store, a store of 123.1 TWh that includes 20% contingency, 89.4 GW of electrolysers, 74% electrolyser efficiency and 55% from hydrogen back to power, a 5% discount rate and wind and solar at GBP 35/MWh (p.30, p.31). No interconnectors (p.19). All electricity comes from wind and solar with hydrogen storage, apart from a small amount needed to regulate voltage and frequency (Figure 13 caption).
Where it comes from
PDF p.18, Table 1 "Weather stress events" (zoom 4 crop), rows as printed: "Up to a week of very low wind speed in winter" with frequency "Every few years"; "Up to four weeks of very low wind speed in summer" with "Once every 10 years"; "One full day of very low wind speed in summer" with "One or two per year". PDF p.17: the table's three types of extreme weather event "were identified in a study based on historical data and modelling" (reference 8, a 2019 Met Office literature review, PDF p.95). PDF p.19: "A study by the Met Office" (reference 13, an unpublished communication) "found that there is approximately a 10% chance per decade of a winter month with wind speeds lower than in the period studied." PDF p.28: "Prolonged periods of low wind, which can occasionally last a few years, risk emptying energy stores if they are not provided with enough contingency." PDF p.31, on Figure 13: a study of the 23 years 1984 to 2006 "would have found a storage volume very much smaller than found by studying" 1980 to 2016, and "there is a very large call on storage in the period" 2009 to 2011 "which reflects persistently low wind speeds". Read from Figure 13 (zoom 4 crop, chart reading): the store's lowest level, about 20 TWh, falls about 274,000 hours into the 37 years, about 31 years in, where the figure marks the contingency. PDF p.6: "Studies based on less than several decades of weather data are liable to very seriously underestimate the need for storage, and overestimate the need for other sources of flexible supply. These under/overestimates are especially large in studies that look only at individual years rather than sequences of years or examine selected periods of high stress." PDF p.78: "The long-term variability of wind creates a need to store tens of TWhs for many years."
Basis: Great Britain. Event frequencies as printed in Table 1 (per year, per 10 years, "Every few years"); store size in TWh of hydrogen (lower heating value). The model uses simulated hourly wind and solar output for the 37 weather years 1980 to 2016 against one 2050 demand profile of 570 TWh a year.
Related records
- Answers question
- QST-021 not published in this release
Cite this
energyaperture.com/c/clm-1488
CLM-1488, Energy Aperture, as of 2023-09.
Connected records
Other published records linked to this one. A list, not a judgement.
- Also from SRC-781
- CLM-1419The Royal Society worked out that a British power system running only on wind and sun, with stored hydrogen to cover calm and dark spells...CLM-1420In the Royal Society's model, only about 15 per cent of Britain's electricity in 2050 comes out of storage, so the price of storage moves...CLM-1469In the Royal Society's model of a British power system run on wind and sun in 2050, electricity fed into the grid costs about 1.7 to 2...
- Cited in articles