A comment piece in the Times earlier this week appeared to make a troubling prediction for the UK wind industry this winter. It warned the arrival of a weather phenomenon known as La Niña, “could cause real problems” for electricity supplies because of the likelihood of it being accompanied by an “unusually cold and windless winter” in the UK.
But, according to Met Office experts and Carbon Brief’s analysis of past La Niña events, there is little ground for such fears. As just one of a host of factors that can affect UK weather, La Niña is unlikely to dominate enough to warrant forecasts of “emergency” supply shortages.
This is especially true this time round since the impending event is expected to be weak, if it materialises at all.
Every five years or so, weakening trade winds cause a shift to warmer than normal ocean temperatures in the eastern equatorial Pacific Ocean, a phenomena known as El Niño. La Niña is El Niño’s cold-water counterpart.
The World Meteorological Organisation this week confirmed there’s a 50-60% chance that a weak La Niña will develop before the end of the year. However, signs that an event may be imminent have been visible for a few months now, fuelling speculation about the likely consequences for everything from weather to food security and commodity prices.
Earlier this week, the Times’ deputy business editor, Robin Pagnamenta, warned that La Niña “may be about to unleash some unwelcome trouble” for the UK wind industry. The claim was amplified by the prominent political commentator Andrew Neil. Pagnamenta said:
This matters, he argued, because of the knock-on effects for how much electricity the UK can generate from wind power. A dip in wind-generated electricity, together with the closure this year of three coal plants and a squeeze on supply from French nuclear plants, means National Grid will be eyeing the winter forecast “more anxiously than ever”, he says.
The Times article echoes a Bloomberg report from May this year, which talks about the potential for “lower wind speeds” during a La Niña and the challenges this could pose to Britain’s power grid. It seems likely the Bloomberg piece is the source of the “windless winter” language used in the Times article. Bloomberg quotes a Deutsche Bank analyst as saying:
So, what’s the evidence for a La Niña-induced “windless” winter in the UK?
Winter in the UK
The last notable La Niña occurred in the winter of 2010/11, with a weak event persisting into the following winter. Before that, the last La Niña was in 2007/8.
La Niña typically peaks between December and April, facilitating a fairly straightforward comparison with wind-generated electricity during the northern hemisphere winter months.
The chart below compares average wind speed in winter for the UK with the electricity output from the UK’s onshore and offshore wind farms for the same period. Meteorologically speaking, winter is defined as December, January and February (DJF).
Average wind speeds are expressed as the deviation from the 10-year average for the three-month period between Dec-Feb (2002-2011). All wind speed figures are in knots (1 knot = 1.151 miles per hour).
Wind-generated electricity is expressed as the average load factor for the same three-month winter period. The chart uses load factors, obtained from data analysts EnAppSys, rather than total generation to account for changes in installed capacity over time. Figures include distributed generation.
Vertical shading indicates the occurrence of a La Niña (blue) or El Niño (red). The darker the shading, the stronger the magnitude of the event.
Average wind speed in winter (Dec, Jan, Feb) in the UK, expressed as the deviation from the 10-year average (purple lines, knots), and the seasonal load factor for offshore and onshore wind-generated electricity (blue columns, %). Vertical shading indicates the occurrence of La Niña (blue) or El Niño (red).
La Niña effect?
For the period with both wind and load factor data (2009-16), there’s a correlation between wind speed and the amount of electricity generated, relative to the maximum theoretically possible. Winters with higher wind speeds are reflected in higher load factors and vice versa.
It’s worth noting that as well as the year-to-year variation with wind speed, there is a general upward trend in load factor over the period 2009-2016 as a result of bigger, more efficient wind turbines being installed. Efficiency gains amount to about a four percentage point increase over this time period, though the incremental effect is more pronounced for offshore because of the smaller size of the fleet relative to onshore.
Looking at the moderate La Niña event in 2010/2011, you can see the average wind speed in winter was higher than 2009/10, but markedly lower than in subsequent years. Averaged over the whole year, 2010 saw the lowest average wind speeds since 2000.
The winter of 2010/11 was also the second coldest in the UK since 1985/86, and followed the exceptionally cold winters of 2008/9 and 2009/10. Freezing conditions resulted in the coldest December in the last 100 years and the coldest across central England since 1890.
The amount of electricity generated from wind in winter 2010/11 reflects the low wind speeds. At 29.6%, the load factor is several percentage points short of the long-term winter average of 35.9%, even after accounting for the small upward trend from efficiency gains.
So, the winter of 2010/11 was less windy than other years, but far from “windless”.
Drilling down into individual winter months reveals some interesting detail. You can see the month-to-month comparison for 2010/11 in the graphs below.
Average wind speed in December (top), January (middle) and February (bottom) in the UK, expressed as the deviation from the 10-year average (purple lines, knots), and load factor for offshore and onshore wind-generated electricity in the same month (blue columns, %). Vertical shading indicates the occurrence of La Niña (blue) or El Niño (red).
During the La Niña winter of 2010/11, average load factors in December and January were below the long-term monthly averages of 38% and 36%, respectively. Wind speeds picked up in February, however, resulting in a load factor of 35.5%, two percentage points above the long-term average for that month.
This signature is quite typical of La Niña, says Prof Adam Scaife, head of the decadal forecasting group at the Met Office. He tells Carbon Brief:
While the winter of 2010/11 might appear, on first glance, to have the signature of La Niña, we should be cautious to interpret patterns in what could just be natural variability.
To investigate this possibility further, the graph below shows the amount of wind-generated electricity on each day of winter, as a percentage of that winter’s average. Deviations above the line show periods of high output, while those below the line represent low output. Colours denote different years. You can switch these on and off to compare output in different years.
During the 2010/11 La Niña winter (red), you can see several short periods when wind-generated electricity dropped to below 5% of the average for that winter. The effect is more pronounced in December, with very few notable troughs occurring in February.
Wind-generated electricity in the UK on each day of winter, expressed as a percentage of that winter’s average (Dec-Feb). Colours represent different years.
The weak La Niña winter of 2011/12 (yellow) shows an entirely different signature, however. Several periods of low output persisted through early February, with December relatively unaffected. That year saw an above average amount of electricity generated from wind over winter, with a load factor of 38.4% compared to the long-term average of 36%.
The moderate La Niña winter of 2007/8 shows a different pattern again, with the third highest average wind speed in the 15-year record. (Unfortunately, no load-factor data is available.)
Importantly, periods of low output are not restricted to La Niña years. Every year has several. It’s worth noting that the longest period of below average output is 17 days between 6-23 Feb during the winter of 2009/10 – an El Niño year. (blue shading in the graph below).
Indeed, if there was a predictable relationship between La Niña/El Niño and wind-generated electricity, as The Times article suggests, you might expect the recent very strong El Niño winter of 2015/16 to show particularly high load factors. At 39.7%, wind generated output in 2015/16 was above average, but not exceptional. More electricity was generated from wind in 2013/14 and 2014/15 (43.3% and 42.1%, respectively), both of which were “neutral” years with neither an El Niño nor a La Niña.
Consecutive days during winter with more than 50% below average wind output, relative to that year’s winter average. Colours represent different years.
No such warnings
All this means that while observations and model studies suggest La Niña can affect European winter weather, other natural fluctuations from month-to-month, or year-to-year, mean the impact of each event is far from predictable. As Scaife tells Carbon Brief:
This would imply little evidence for the assertion in the Times article that an impending La Niña is likely to reduce wind-generated electricity this winter to such an extent that it will “force National Grid to draw on emergency back-up supplies”. This is not to say we won’t see periods of low output, but if we do it will be hard to pin the blame on La Niña.
Indeed, National Grid is issuing no such warning. Its Winter Outlook, released last week, said:
“We expect there to be sufficient generation and interconnector imports to meet demand throughout winter 2016/17. The electricity margin is similar to last year but includes a larger proportion of contingency balancing reserve services.”
Finally, scientists expect the coming La Niña to be a weak one, so the impacts are unlikely to dominate the UK’s notoriously complicated weather. With about half of climate models predicting a return to “neutral” conditions, there’s even a fair chance the much-hyped La Niña won’t materialise at all.