I drew attention to the recent Oxford study that claimed we could have an all wind and solar energy supply.
I went digging, via the OU press release to a departmental announcement that was a draft for the press release to their 3 page (really barely 1 page) summary for policy makers and finally to their working paper, whose main intent is expressed as undermining David Mackay's SEWTHA which is transparent in its assumptions and at least uses correct physics.
It's models all the way down. They refer to BEIS estimates of costs that have been completely debunked by the failed AR5 CFD auction. This is what they had to say about their assumptions for wind turbine power curves in footnote 14:
14 Power at a given wind speed (𝑃𝑔) was calculated using function 𝑔(𝑣), where the shape of 𝑔 varies by turbine but is broadly
defined as 0 below cut-in wind speed, some value between 0 and the maximum (𝑃𝑔𝑚𝑎𝑥 when 𝑣 is between cut-in speed and
rated speed, 𝑃𝑔𝑚𝑎𝑥 when 𝑣 is between rated speed and cut-out speed, and 0 when 𝑣 is above the cut-out speed.
That is they refuse to show their assumption or detail how they derived them. Instead of looking at real world historical data on actual wind speeds they resort to a model:
13 This was computed using a probability density function of typical two-factor Weibull form:
at say £1.25m/km that's a mere £100 trillion bill for the world to pay. There may be an awful lot of coffee in Brazil, but is there enough copper in Chile?
When I do these analyses I always consider the range between an all solar and an all wind solution, varying e.g. the solar from 100% down to 0% in tranches of say 20% initially. This allows some insight on how the major variables influence costs for grid capacity and storage etc..and also to see whether there is a reasonably flat optimum that might allow some flexibility to handle changes in relative costs. The RS really should have presented such an analysis explicitly up front, instead of the somewhat confused discussion in the supplementary appendices.
The next stage is to examine the trade off between storage and overbuild/curtailment, and indeed to look at how the storage/curtailment requirement escalates on the path towards an all renewables grid. That would reveal that costs will start escalating alarmingly once we go beyond about 60% of energy supplied by renewables. I have provided analysis of that to OFGEM and BEIS and the BEIS Select Committee that appears so far to have been studiously ignored.
Of course, they also failed to consider the cost landscape adequately, and the real world performance of wind turbines. I am thinking of writing to Iain Staffell and asking whether there are plans to make available the RS wind and solar datasets at his renewables.ninja website, as has been done for some other papers that relied on his weather reanalysis tools. Then we can get a much better handle on the extent of rosiness in the workings.
I drew attention to the recent Oxford study that claimed we could have an all wind and solar energy supply.
I went digging, via the OU press release to a departmental announcement that was a draft for the press release to their 3 page (really barely 1 page) summary for policy makers and finally to their working paper, whose main intent is expressed as undermining David Mackay's SEWTHA which is transparent in its assumptions and at least uses correct physics.
It's models all the way down. They refer to BEIS estimates of costs that have been completely debunked by the failed AR5 CFD auction. This is what they had to say about their assumptions for wind turbine power curves in footnote 14:
14 Power at a given wind speed (𝑃𝑔) was calculated using function 𝑔(𝑣), where the shape of 𝑔 varies by turbine but is broadly
defined as 0 below cut-in wind speed, some value between 0 and the maximum (𝑃𝑔𝑚𝑎𝑥 when 𝑣 is between cut-in speed and
rated speed, 𝑃𝑔𝑚𝑎𝑥 when 𝑣 is between rated speed and cut-out speed, and 0 when 𝑣 is above the cut-out speed.
That is they refuse to show their assumption or detail how they derived them. Instead of looking at real world historical data on actual wind speeds they resort to a model:
13 This was computed using a probability density function of typical two-factor Weibull form:
𝑓(𝑣) =𝑘/𝑐 ∙ (𝑣/𝑐)^(k-1)∙ 𝑒^−(𝑣/𝑐)^𝑘, 𝑐 > 0, 𝑘 > 0, 𝑣 > 0,
where 𝑣 is wind speed, 𝑐 is the scale parameter, and 𝑘 is the shape parameter.
And on and on. It's worthless.
Another Net Zero bill. According to the IEA there is a need to install 80 million km of transmission lines
https://www.energylivenews.com/2023/10/17/nearly-80-million-kilometres-of-grids-needed-by-2040-to-meet-climate-targets/
at say £1.25m/km that's a mere £100 trillion bill for the world to pay. There may be an awful lot of coffee in Brazil, but is there enough copper in Chile?
When I do these analyses I always consider the range between an all solar and an all wind solution, varying e.g. the solar from 100% down to 0% in tranches of say 20% initially. This allows some insight on how the major variables influence costs for grid capacity and storage etc..and also to see whether there is a reasonably flat optimum that might allow some flexibility to handle changes in relative costs. The RS really should have presented such an analysis explicitly up front, instead of the somewhat confused discussion in the supplementary appendices.
The next stage is to examine the trade off between storage and overbuild/curtailment, and indeed to look at how the storage/curtailment requirement escalates on the path towards an all renewables grid. That would reveal that costs will start escalating alarmingly once we go beyond about 60% of energy supplied by renewables. I have provided analysis of that to OFGEM and BEIS and the BEIS Select Committee that appears so far to have been studiously ignored.
Of course, they also failed to consider the cost landscape adequately, and the real world performance of wind turbines. I am thinking of writing to Iain Staffell and asking whether there are plans to make available the RS wind and solar datasets at his renewables.ninja website, as has been done for some other papers that relied on his weather reanalysis tools. Then we can get a much better handle on the extent of rosiness in the workings.