Sunday
May262013
by 
Bishop Hill
Bishop Hill Ian Fells on energy shortfalls
May 26, 2013 Energy: gas Energy: grid Professor Ian Fells was interviewed on the BBC's PM programme about the story that Britain recently came within 6 hours of running out of gas. He is very interesting on the electricity supply problems a couple of years ago, suggesting that, despite official denials, these were due to a lack of gas.
The audio file is attached below.
Reader Comments (27)
Anyone who lives in a major city is just 12 hours away from anarchy!
See James Burke's 30 year old documentary about the 1965 blackout in New York.
There are a number of stories circulating around the theme of "Britain within 6 hours of running out of gas."
Britain has a number of strategically located Gas Storage facilities such as at the Isle Of Grain, Rough (a depleted gas field in the North Sea), Hornsea, Avonmouth etc.
Supplies from these storage facilities were virtually depleted, mainly due to the prolonged cold spell (i.e. it was the duration rather than the severity).
In addition to 'Storage', our Natural Gas Transmission System is such that the distribution pipelines contain about 1 - 2 day's worth of supplies en-route to users.
National Grid (and British Gas before them) have always had specific procedures for a worse than 1-in-25 winter scenario (or serious localised problem).
'Interruptibles' (very large industrial/commercial customers who pay less, knowing that their gas supplies will be cut-off for a number of days each year) are always the first to be load-shed.
In an emergency, further loads are shed by instructing medium & larger businesses using gas for non-critical purposes (e.g space heating) to shut down. (i.e to "Slim" the system.)
Finally, to "Sever" the system, the remaining large businesses (Industrial process users, Hospitals, Power Stations on 'Firm' contracts) are instructed to shut down.
All so that the distribution system remains pressurised, and can meet domestic demand.
The correct headline for the original scenario was "Britain was only SIX HOURS away from the lights being switched off in March as gas supplies ran 'dangerously low' " - i.e supplies to power stations being forcibly interrupted.
I venture that everyone in the UK has experienced electricity power cuts multiple times - we had two brief interruptions two days ago; very few domestic consumers have ever had a district gas interruption.
This has got to be good news for village fairs, potters, bakers and blacksmiths.
Forget about stocks and shares; invest in rotten fruit and ploughshares !
Oh and don't forget to block that Chunnel thing; as we enter the New, improve Middle-Aged' we don't want History to repeat itself.
On a happier note waterboarding may be replaced by serfboarding and those unsightly turbines will be replaced by picturesque windmills.
If the Met Office concentrated on its prime function of accurate and timely forecasting of the UK's weather over various timescales, crises such as the above would rarely occur.
There is a case for banning the MO from any guesswork into future 'climate'. I certainly begrudge my taxes contributing towards their profligacy.
Joe Public
There is no sharp cutoff point between "accurate and timely forecasting" and "guesswork into future climate". You can produce a very accurate forecast for six hours ahead, accurate forecasts for a day ahead and a general forecast for 5 days. Beyond that, accuracy falls off rapidly, to the extent that the Met Office no longer publishes seasonal forecasts. Over climate timescales of 30 yers plus, all that is possible is estimates of long term trends.
Since meteorology studies a chaotic process, this is all you can get. As you timescale increases detailed forecasts become more difficult until only the long-term trends remain predictable.
Is it entirely a coincidence that the accuracy of met forecasts decays with the ability via satellite and observation to see what is coming from somewhere else? In fact they don't really predict much at all other than that.
Entropic Man:
"...only the long-term trends remain predictable."
While I agree with you that it remains beyond our reach to forecast details beyond a few days (at the moment), the long-term trends are *not* currently predictable. At a global level, the average surface temperature trend has been overpredicted by the CMIP3 and now CMIP5 model ensembles, and precipitation trends are less accurate. At the regional level, the picture is even worse. I am optimistic that the situation will improve, but it is incorrect to state that the long-term trends are predictable now.
Energy suppliers held back gas during UK shortage: http://www.guardian.co.uk/business/2013/may/24/energy-suppliers-held-back-gas-uk
Just somehting to manipulate the prices?
"...only the long-term trends remain predictable."
I predict that on the long term I will win the lotery...
HaroldW
My main point was that accuracy decreases with timespan. The 95% confidence limits which should accompany any prediction get much larger as you look further ahead for just that reason.
The universe also has a way of changing things unexpectedly. I doubt that CMIP3 included the weak maximum for solar cycle 25. Prediction is only as good as the information and understanding available at the time.
EM, my point is that when the Met can actually see the weather coming they can forecast it. This does not seem to me to be a very big deal, any more than when one sees a dark cloud upwind and gets the washing in off the line. It is plain that when they can't see it coming they are no more likely to get the forecast right than anybody's educated guess. There's no shame in that, but neither is there any grounds for boasting about their forecasting prowess or their modelling ability.
A transcript of the Ian Fells interview is now available here:
https://sites.google.com/site/mytranscriptbox/home/20130524_pm
I presume in basic terms the dangers in starting up domestic supplies if they were to be cut off is the risk of getting air in the pipes, creating an explosive mix? Any technical input on that?
@ mike fowle at 3:45 PM
"... risk of getting air in the pipes, creating an explosive mix?"
Nearly correct. That, and the need to have every affected pipe fully purged before gas supplies can be reinstated.
Any fire or cooker without flame-failure device extinguishes, but is not physically turned off. So on resumption of supplies, neat gas would be emitted. That, combined with atmospheric air produces an explosive situation, waiting to be triggered by the next electrical appliance that is switched on.
It is for this reason that the gas industry goes to such enormous lengths to maintain continuity of supply.
"Since meteorology studies a chaotic process, this is all you can get. As you timescale increases detailed forecasts become more difficult until only the long-term trends remain predictable."
I think Entropic Man is sticking his neck out here. The advantage is that "long-term" is long enough for his comment to be forgotten. There is simply no current evidence that supports the idea that "long-term trends remain predictable".
May 26, 2013 at 3:24 PM | Rhoda
Exactly, I was studying at the same time the ECMWF started their ensemble forecasts in the early nineties. They ran 16 models with the observations perturbed to maximise divergent forecasts. Just by glancing at the 16 forecasts you could see a high pressure dominated forecast was very predictable while typical UK depressions tracking in off the Atlantic diverged significantly within a day. Hence typical UK weather can't be forecast very well without a bit of lucky guessing.
Thanks Joe Public re the risks of cutting off supplies.
EM
Since meteorology studies a chaotic process, this is all you can get. As you timescale increases detailed forecasts become more difficult until only the long-term trends remain predictable
and
The universe also has a way of changing things unexpectedly. I doubt that CMIP3 included the weak maximum for solar cycle 25. Prediction is only as good as the information and understanding available at the time.
So how can the first statement be true when the unknown unknown in the second could well be one of a huge number? Even the known unknowns are a bit of a problem are they not?
May 26, 2013 at 6:40 PM | SandyS
Both statements are true on Planet MetOffice.
Billy Liar
You're probably right.
Two points.
1. Removal of gas for space heating will tend to raise electricity consumption but there is a serious additional issue: electricity generation using gas has a much lower thermal efficiency than gas space heating. Any gas fired power stations must not come on line or increase output. Remind me which stations are used for rapid response and how their thermal efficiency varies over time?
2. I would put money on demand being the same as weather, does not follow a a simplistic random law as assumed by many and with statistics so assuming. The Hurst exponent is elevated hence the "periods of" abnormal weather except it isn't, it's normal.
This requires altered statistical formula and leads to considerably greater variability. Here is how come there are declarations of 100 year or 1000 year exceptions in rapid succession.
I demonstrated complete with grab a tool that eg. CET has a Hurst exponent around 0.7, not the 0.5 of gaussian. (do not use annual data, too short, are looking for self similarity at different scales)
Rhoda, SandyS
Among the unknown unknowns is whether you will be alive tomorrow. The implication of your philosophy is that the future is entirely unknowable and any planning for the future a waste of time and effort.
Among the unknown unknowns is whether you will be alive tomorrow. The implication of your philosophy is that the future is entirely unknowable and any planning for the future a waste of time and effort.
I might indeed die tomorrow. Indeed there must be a day when the day after I will be dead. So I plan for that situation (I have a will, for example) and also for the alternative. Unpredictability is no deterrent to planning.
In any case, just because long range weather is more or less entirely unpredictable doesn't mean other things are.
[snip] I will spell out the implications. Greens tell us CO2 is warming the earth and we must drastically reduce our lifestyle because of an unpredictable future. Sensible people say, no, we prefer to mitigate thanks, since that way we can hedge our bets on the unpredictable.
Some people like to feel confident that they can predict the future. That makes them feel good, but in fact that feeling is a false sensation. Ones confidence in predicting the future is almost always unrelated to your actual ability to predict it. Bookies make a good living on the fact that people are almost always overconfident about their ability to predict.
Wishful thinking and self delusion, on your part and on the Met Office's part.
Although the Met Office is immensely more culpable for the consequences of its bullshitting than is someone who has simply been convinced by their propaganda. After all, how could what comes out of a 100 squillion megaflop computer possibly be wrong? Especially when it has been "validated" be reproducing the past history that was available to the Met Office programmers?
The inability of the Met Office's models (their - inherently - unvalidated models) to predict even medium term trends is plain for everyone to see. Not a big surprise to anyone who has spent time modelling physical systems.
"Among the unknown unknowns is whether you will be alive tomorrow. The implication of your philosophy is that the future is entirely unknowable and any planning for the future a waste of time and effort."
You deal with all this based on common sense. When considering whether you will die tomorrow you deal in likelihoods. Shall I spend all my money and health having fun, or should I plan for every contingency and save everything only to get too old to do anything with things left undone? That's the human condition, EM, and you have discovered nothing when you see the false implication you mention. In the case of the weather, it is not entirely unknowable, you can be fairly confident we will see weather and climate within observed limits. I'm not aware of any recent weather trumpeted as extreme which actually has set any new records. And that is the assumption I will work on. I am planning for the possibility of rain today, but not a rain of fiery meteorites.
EM
Which is worse planning for a single eventuality or planning taking into account that we don't know what the future holds?
You seem to favour the former experience has taught me, i can't speak for Rhoda, that the latter is what I should do. Cross that bridge when we get to it if you like. Good things (unknown unknowns) might happen to make crossing the bridge easier, why be pessimistic about the future?
I work in the energy industry and Fells is right. A few years ago there were many projects in development to increase our current gas storgare capability of about 5 days to that of other EU countries. France and Germany have about 100 days storage capacity. The facilities would have been in exhausted oil wells offshore or in underground salt caverns.
However there were three big issues at the time. The taxation paybale on "cushion gas" which essentially stays in the storage site and also the Crown increased fees for use of the sea bed. Thirdly the Labour government at the time were going feet first into renewables i.e. wind. These reasons made the developements uneconomic for the energy companies. Whilst a few did get off the blocks most were dropped. So we remain with very little storage capacity and in worst case scenario we could lose our gas supply.