Controversial new study on wind energy with findings also misrepresented by Elsevier
Robbie Morrison
Hello all
Caveat: the underpinning PDF is paywalled so this posting is based on the article's public highlights and summary (in lieu of an abstract), a New Scientist report (Page 2018), and an associated press release from academic publisher Cell Press (2018).
A new study by Miller and Keith (2018) from Harvard University claims that if the continental United States decided to install sufficient wind turbines to meet current electricity demand, one third of the surface area would be covered by wind farms and surface temperatures would rise on average by 0.24°C as a result. The authors then compare this increase in temperature with the effects of global warming.
The ‘warming’ from
wind turbines is due to their mixing of stratified air near
ground level, where the land surface is typically colder,
particularly at night. This effect can, unlike global warming,
be halted in an instant by simply curtailing generation.
Cell Press, owned by Elsevier and publisher of the journal Joule, issued a press release headed “Large-scale US wind power would cause warming that would take roughly a century to offset” (Cell Press 2018). The release goes on to imply that the savings in greenhouse gas emissions from these wind turbines would not counteract their warming effect for about 100 years. Quoting from the press release:
”The direct climate impacts of wind power are instant, while the benefits accumulate slowly,” says Keith. “If your perspective is the next 10 years, wind power actually has – in some respects — more climate impact than coal or gas. If your perspective is the next thousand years, then wind power is enormously cleaner than coal or gas.”
New Scientist opted to run a scathing commentary on the Miller and Keith paper (Page 2018). Issues raised include the improbability of wind deployment on this scale, the degree to which wind turbines actually mix air (possibly overstated by a factor of three), and the conflating of mixing and warming processes. New Scientist is clearly of the view that the Cell Press statement is intellectually misleading and the article itself is deeply flawed.
Wikipedia has an article on Keith and reports he is a proponent of geoengineering, including the spraying of reflective particles in the upper atmosphere.
Miller and Keith (2018b) have also just published a related open access paper titled “Observation-based solar and wind power capacity factors and power densities”. I read the abstract and it seems considerably more sensible. Power densities are however a second‑tier consideration at best and should not be accorded much more than a footnote in relation to public policy development. The tier‑one considerations are carbon, cost, and security, with environmental impacts best handled directly and not by some proxy metric, in this case power density.
(When I studied
engineering, we learnt that fossil fuels had just the right
energy density, with renewable too low and nuclear too high.
Also that R‑12 was cheap, safe, and nontoxic.)
I rather think we are
going to see more controversies of this nature going forward.
They will doubtless vary in merit, grounding, and status. In any
case, best not to dig bunkers (I am thinking of climategate
here) but rather to be as open, clear, and reproducible as
possible. On that note, does anyone know if the Miller and Keith
code and data is available and lawfully runnable?
with best wishes, Robbie
PS: Email me if you
would like a copy of Page (2018).
Note that sub‑sections §60c(1) and (3) of the German copyright act (Juris 2018) allow me to do so under the following provisions:
(1) Up to 15 per cent of a work may be reproduced, distributed and made available to the public for the purpose of non-commercial scientific research 1. for a specifically limited circle of persons for their personal scientific research and 2. for individual third persons insofar as this serves the monitoring of the quality of scientific research.
(3) In derogation from subsections (1) and (2), full use may be made of illustrations, isolated articles from the same professional or scientific journal, other small-scale works and out-of-commerce works.
If anybody has access
to the Miller and Keith article, can they offer to circulate it
on request too? Also send me a copy. TIA.
References
Cell Press (4 October 2018). Large-scale US wind power would cause warming that would take roughly a century to offset. Science Daily. USA.
Juris (2018). Act on Copyright and Related Rights (Urheberrechtsgesetz, UrhG) — Amendments to 1 September 2017 — Official translation. Saarbrücken, Germany: Juris.
Miller, Lee M and David W Keith (4 October 2018). “Climatic impacts of wind power”. Joule. ISSN 2542-4351. doi:10.1016/j.joule.2018.09.009. Paywalled.
Miller, Lee M and David W Keith (2018b). “Observation-based solar and wind power capacity factors and power densities”. Environmental Research Letters. 13 (10): 104008. ISSN 1748-9326. Open access.
Page, Michael Le (13 October 2018). “Wind power’s warming effect is overblown”. New Scientist. (3199): 25. ISSN 0262-4079. The online title differs thus: “Wind farms do affect climate — but they don’t cause global warming”. Paywalled.
-- Robbie Morrison Address: Schillerstrasse 85, 10627 Berlin, Germany Phone: +49.30.612-87617
Robbie Morrison
Hello again
Here is a link to the article in question:
--
You received this message because you are subscribed to the Google Groups "openmod initiative" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openmod-initiat...@googlegroups.com.
To post to this group, send email to openmod-i...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/0ffac534-ea21-8a95-9285-ff976df886c1%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Jan Wohland
Dear Robbie, all,
thanks for this posting. I have two remarks on the context:
There is a recent publication by Pryor and colleagues on this
topic which concludes that local climatic effects of wind park are
small [1]. The difference might simply be the installed capacity
but I assume that there must be more fundamental differences in
approaches. There also was a similar study for Europe a couple of
years ago which also concluded that effects are in deed negligle
[2].
With respect to the temperature argument, I am not convinced
either. Global mean temperature currently rises by around 0.2
degrees/decade [4] and the increase is generally stronger over
land. It would thus take something around a decade to reach 0.24
degrees of global warming over land. It is by far overstated to
talk about milennia here (in fact there are two orders of
magnitude in between).
Best
Jan
[1] Pryor, S. C., Barthelmie, R. J., & Shepherd, T. J. (2018). The influence of real‐world wind turbine deployments on local to mesoscale climate. Journal of Geophysical Research: Atmospheres, 123, 5804–5826. https://doi.org/10.1029/2017JD028114
[2] Vautard, R., Thais, F., Tobin, I., Bréon, F.-M., de Lavergne, J.-G.D., Colette, A., Yiou, P., Ruti, P.M., 2014. Regional climate model simulations indicate limited climatic impacts by operational and planned European wind farms. Nature Communications 5. https://doi.org/10.1038/ncomms4196
[3] Myles Allen et al., IPCC Special Report on 1.5, Technical
Summary, http://report.ipcc.ch/sr15/pdf/sr15_ts.pdf
--
Johannes Schmidt
Peter Jordan-Str. 82
1190 Vienna / Austria
Robbie Morrison
Hello Johannes, all
Thanks for your reply. My reading is that Miller and Keith (in their abstract at least, see later) use their power density metric in a quite different manner from the one you describe. Instead of referring to some average installed power density as an interpretation that can later be translated into turbines per unit area for instance, as you indicate, Miller and Keith use the metric to characterize low carbon generation technologies with the idea that technologies with low innate power densities should be discriminated against by public policy on the grounds of presumed higher environmental impact. While that line of argument may be useful for running down wind power (note that wind power is normalized by wind farm extent and not installation footprint) vis‑a‑vis photovoltaics — the not-so-hidden agenda of the authors I would say — I don’t really see any merit in this approach otherwise.
Better to undertake sufficiently detailed numerical studies and explicitly account for specific and known environmental impacts by some or all of the following methods:- scope-limited scenarios (such as respecting protected areas)
- run-time wildlife protection (such as curtailing wind farms in misty weather during migration times when birds in transit are known to fly closer to the ground or sea surface)
- hard environmental constraints (such as CO2 emissions ceilings)
- optimization penalties on environmental transgressions (penalty methods and similar)
- multi-criteria optimization (for completeness, no energy model I know of supports this feature)
- report aggregate and peak impacts for later assessment (human decision makers are generally good at factoring in secondary criteria)
- favor pareto over non-pareto solutions in the absence of other considerations (examine only the best performing scenarios)
That said, the assumption by (Jacobson et al 2018) that urban rooftops can be retrofitted with a capacity density 4.5‑times higher than the commercial-scale solar plants measured by (Ong et al 2013) seems highly unlikely, as does the resulting 24–27 We m−2 power density (Jacobson et al 2018).
Of course, Jacobson looks forward towards 2050 and the measured data relates to legacy plant of unstated vintage (I didn't check Ong et al so maybe those details are covered). Anyway that is another reasonable use of scale-independent (also known as intensive) metrics: integrity checking to flag potentially spurious input or suspect output. On that note, anyone undertaking renewables assessments should consider checking their metrics against the values in this paper. Interpreted metrics are thus useful, but labeling technologies instead of undertaking integrated analysis is poor practice. On that exact note, I am relieved that the term “base load plant” has now largely been forgotten.
Here is a quote pertaining to Germany (p9):
As an example of the implications of these results, consider Germany and its ambitious energy transformation policy (Energiewende). Germany’s primary energy consumption rate is 1.28 Wm−2 (BP 2018). If our US wind power density of 0.50 Wem−2 [e = electrical] was applicable to Germany, then devoting all German land to wind power would meet about 40% of Germany’s total primary energy consumption, while if German wind power performs like the best 10% of US wind (0.80 Wem−2), then generation would be 62% of Germany’s consumption. Finally, if Germany’s goal was to generate the most wind power without economic constraints, very high capacity densities (e.g. 10 MWikm−2) [i = installed] could be deployed, reducing capacity factors but possibly raising the power density to 1.0 Wem−2 and meeting 80% of consumption. Whereas for solar at 5.4 Wem−2, 24% of Germany’s land area would need to be devoted to commercial- scale solar to meet total primary energy consumption.
German modelers — is
this okay then? Clearly a mistake to equate coal to electricity in
primary energy terms (or just maybe the BP document reports exergy but then
an "e" subscript should be present on the "W").
What is really odd
about this paper is the abstract mentions environmental
consequences in relation to power density three times,
but the issue is never covered in the body of the paper nor even
raised in the discussion. And what exact environmental — as
opposed to physical — consequences are we talking about? What
exactly does a modest localized surface temperature rise (circa
0.2°C on average according their other paper) mean in ecological
or economic terms? (Surely its time to move to open peer review.)
The other Miller and Keith paper on the climatic impacts of wind power made it into one major German daily (Grolle 2018) — paywalled so I have not seen more than the title. It may well be that the Spiegel article was generally critical although with a click-bait headline.
Finally, here is a
really odd but nonetheless carefully crafted video (02:39) with
Miller explaining the paper:
Miller conflates mixing
and heating and micro‑climate with global climate. The video then
a depicts a photovoltaic installation shading land with Miller
claiming that the [micro-]climate impacts are negligible. I can
only class these two papers as advocacy research.
with best wishes, Robbie
References
Grolle, Johann (9 October 2018). Windenergie: Droht das Ende der Wind-Ernte? [Wind energy: the end of the wind harvest threatened?] (in German). Spiegel Online. Hamburg, Germany. Paywalled.
Jacobson, Mark Z, Mary A Cameron, Eleanor M Hennessy, Ivalin Petkov, Clayton B Meyer, Tanvi K Gambhir, Amanda T Maki, Katherine Pfleeger, Hailey Clonts, Avery L McEvoy, Matthew L Miccioli, Anna-Katharina von Krauland Rebecca W Fang, and Mark A Delucchi (1 October 2018). “100% clean and renewable Wind, Water, and Sunlight (WWS) all-sector energy roadmaps for 53 towns and cities in North America”. Sustainable Cities and Society. 42: 22–37. ISSN 2210-6707 : tags. doi:10.1016/j.scs.2018.06.031. Paywalled.
Miller, Lee M and David W Keith (2018). “Observation-based solar and wind power capacity factors and power densities”. Environmental Research Letters. 13 (10): 104008. ISSN 1748-9326. Open access.
Miller, Lee M and David W Keith (4 October 2018). “Climatic impacts of wind power”. Joule. ISSN 2542-4351. doi:10.1016/j.joule.2018.09.009.
Tröndle Tim
Renewable Energy Policy Group
Universitätstrasse 22, CHN J72.1
8092 Zürich, Switzerland
email: tim.tr...@usys.ethz.ch
office: (+41) 44 632 28 08
web: www.rep.ethz.ch
--
You received this message because you are subscribed to the Google Groups "openmod initiative" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openmod-initiat...@googlegroups.com.
To post to this group, send email to openmod-i...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/fefedca1-2789-c782-8a90-6a56c3856db8%40posteo.de.
Robbie Morrison
Hello again all
I tried to track back from the 1.28 Wm−2 value which represents “Germany’s primary energy consumption rate”. The source quoted is BP (2018). Let’s assign this metric to G.
First, no mention of exergy or second‑law analysis in BP (2018) so that issue can be ruled out straightaway (I have previously argued that national energy accounts would be better expressed using exergy).
Primary energy consumption rate
Annual primary energy consumption for Germany in 2017 as reported in BP (2018:8):
- 335.1 Mtoe (million tonnes oil equivalent)
And wikipedia states the land area of Germany is:
- 357 386 km2
and duly references an official source which confirms that value.
Calculations as follows (the GNU units utility is pretty useful at times):
$ units --verbose --one-line --digits=3 "335 Mtoe" "J"
335 Mtoe = 1.4e+19 J
$ units --verbose --one-line --digits=3 "1 year" "s"
1 year = 3.16e+07 s
$ units --verbose --one-line --digits=3 "357386 km^2" "m^2"
357386 km^2 = 3.57e+11 m^2
G = 1.4e+19 J / 3.16e+07 s / 3.57e+11 m^2 = 1.24 Wm^2
Not far off — within 3% of — the Miller and Keith figure but still a little outside round‑off errors based on the quoted input. In any case, let’s accept the original figure.
Accounting for thermal conversion
Thermal generation first-law efficiencies typically range from 30–60% under on-design conditions. BP (2018:9) reports the following primary energy breakdown for Germany in 2017:
| Fuel | Mtoe |
|---|---|
| oil | 119.8 |
| natural gas | 77.5 |
| coal | 71.3 |
| nuclear | 17.2 |
| hydroelectricity | 4.5 |
| renewables | 44.8 |
| total | 335.1 |
(119.8 + 77.5 + 71.3) / 335.1 = 80%
17.2 / 335.1 = 5%
So as a rough estimate (not my favorite type of analysis), 80% of the primary energy mix has an average conversion efficiency of say generously 45%, while the remainder is accounted in terms of electricity (see note below) hence, the overall primary energy to electricity efficiency is as follows:
0.80 * 0.45 + 0.20 * 1.00 = 56%
I understand that the nuclear figures are for “gross generation” and have already been adjusted using an average conversion efficiency of 38% (BP 2018:41).
The above calculation means that G is overstated in the paper by around two-fold: clearly a serious methodological error.
Note also that a fully electrified system can use heat-pumps running COPs well above unity to supply some of the low grade thermal energy currently serviced by conventional boilers.
But this whole exercise is almost completely pointless. Covering the entire land-mass of Germany with wind turbines is meaningless. And there are there a number of detailed numerical studies that look at 100% renewables scenarios for Germany — some admittedly confined to only final electricity demand, albeit with increasing shares as electrification proceeds. Wikipedia lists some of those studies. These studies and others should have been reviewed in the paper.
If anyone thinks a
rebuttal to both papers is in order, I am happy to contribute to
a critique of the power density article.
with best wishes, Robbie
References
BP (June 2018). BP statistical review of world energy 2018 (67th ed). London, United Kingdom: BP.
--
You received this message because you are subscribed to the Google Groups "openmod initiative" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openmod-initiat...@googlegroups.com.
To post to this group, send email to openmod-i...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/fefedca1-2789-c782-8a90-6a56c3856db8%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
-- Robbie Morrison Address: Schillerstrasse 85, 10627 Berlin, Germany Phone: +49.30.612-87617
Robbie Morrison
Hi Tim, all
“Note that the amount of primary energy required to supply the same amount of final energy will fall with electrification and battery storage-reducing requirements, but using electricity to make gas or other synthetic fuels has the opposing tendency.” (p. 9)
That statement is actually pretty meaningless too. What exactly are "battery storage‑reducing requirements": is that a reference to demand‑side response or long‑range transport and trading or what? The overall conversion efficiency depends entirely on which technologies are deployed and how well they coordinate (a future AfD‑led government might well stall or reverse the coal phase‑out while electrification continues apace?). Moreover, intertemporal constraints were entirely absent in the original "analysis", so including any battery storage will make the interpretation worse — as is correctly indicated for power-to-X architectures.
Why do this stuff when far better methods and completed studies
exist? As Tim indicates, the historical analysis is useful and
interesting, but using these various metrics to inform future
system architectures and drive public policies is not.
Robbie
Robbie Morrison
Hello Tom, all
Tom Brown kindly forwarded me some rebuttals from Mark Jacobson (WWS project, Stanford University) and Michael Goggin (Grid Strategies LLC, Washington DC, USA). Both respondents address both Miller and Keith papers.
Power densities
Mark challenges an average United States solar power output density of circa 5.0 Wem−2 and responds that real power densities are in the order of 30–40 Wem−2 for rooftop and canopy PV and 17–22 Wem−2 for utility PV. Mark cites trade information to back up these claims.
Both Mark and Michael challenge an average United States onshore wind power output density of 0.5 Wem−2. Michael quotes a National Renewable Energy Laboratory figure of 2.9 Wem−2 (p4 with no source cited). Whereas Mark responds that 4.0–8.5 Wem−2 is more correct. Michael also says (p4) that a value of 3.0 Wem−2 is regularly used by the US Department of Energy (DOE) and also other analysts.
Both Mark and Michael criticize the polygon method for establishing wind farm extent, but both fail to note that Miller and Keith use the median Voronoi cell area to determine the overall size — this then being multiplied by the number of turbines. The method needs further verification, but on first glance, it looks reasonable to me. However, Mark points out that wind turbines may be spaced out using considerations other then optimal harvest, thereby yielding distorted, or at least suboptimal, values. On the other hand, capacity factors are not much influenced by looser spacings. Mark has a paper under review on these issues (Enevoldsen and Jacobson).
How could the values in the original paper clear peer review when evidently out by an order of magnitude, give or take? Have analysts or modelers in Germany calculated these metrics for their datasets and/or simulations? That would be really interesting to know and compare with the Miller and Keith results.
Both respondents challenge the available land area methods and estimates for wind generation as used by Miller and Keith and also their resulting resource usage in light of their low average power density estimates.
Wind farm warming effects
Michael challenges the merits of equating the temperature effects from highly localized wind turbine mixing with long‑haul climate change. Michael also adds that night time warming — which is dominant — should be far less detrimental for agriculture than day time warming and indeed some studies show (again not cited) that wind farms can beneficially reduce summer peak temperatures.
In contrast, Mark say that wind turbines contribute to a reduction in global water vapor — a significant greenhouse gas — by slowing evaporation and that Miller and Keith fail to consider this beneficial effect on the global climate. Mark further states that this mechanism is the main way in which wind turbines directly affect global temperature change.
Mark therefore opines (p1) that the Miller and Keith “results are 100% wrong and should not be relied on to affect policy in any way”.
Closure
I should add that Mark is no stranger to scientific controversy and his claims, like any other, need to be corroborated or alternatively be seen as outliers. That said, modelers in particular need to estimate future parameters in some manner and there can be no correct answer.
It should be noted
that Mark cited only his own work, but that may well have been
due to time constraints.
It is my firm belief that well-executed representative numerical simulations can avoid this seemingly unresolvable debate that characteristic metrics and labeling more generally seem to engender. For comparison, it was okay to analyze pipe flow using the Manning formula before CFD but far better techniques now exist and no one needs to argue over the parameterization and scope of that now obsolete formula (actually when tested more recently against CFD simulations, the original formula was surprising well honed — those old engineers were clearly smart).
with best wishes, Robbie
References
Enevoldsen, Peter and Mark Z Jacobson (under review). “Data investigation of installed and output power densities of onshore and offshore wind turbines worldwide”.
Goggin, Michael (5 October 2018). New studies cause confusion about benefits of renewable energy. Washington DC, USA: Grid Strategies LLC.
Jacobson, Mark Z (2 October 2018). Response to Miller and Keith “Climatic Impacts of Windpower” (Joule, 2018).
Jacobson, Mark Z (3 October 2018). Response to Miller and Keith “Observation-based solar and wind power capacity factors and power densities” (Environmental Research Letters, 2018).
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/d23c2cdd-a4a1-901a-5d13-6a141778457d%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Robbie Morrison
Hello all (cc: Mark Jacobson and Peter Enevoldsen FYI)
The story continues. I emailed Mark Jacobson and he gave permission to cross-post his reply (which is below mine).
But first some comments from me. Mark issued another rebuttal (Jacobson 6 October 2018) following further dialog with the original authors Lee Miller and David Keith (hereafter "M+K"). That PDF has numbered statements in bold from the Harvard researchers and Mark's responses set in normal weight following.
Taking the power density debate first. Mark still regards the wind figures as 16-fold too low and the PV figures as 3 to 4-fold too low. Moreover, Mark claims there are errors in the use of other authors' results.
Issues 1 thru 3 traverses the use of Voronoi polygons to estimate
wind farm areas. M+K describe their method. In contrast, Mark
concentrates on the values produced. Mark also cites Enevoldsen
and Jacobson in support which M+K point out has not cleared peer
review. Small wind farms are apparently difficult to assess, what
ever method is used.
Issue 4 turns to PV. M+K say they primarily relied on Ong et
al (2013). In response, Mark criticisms the methods employed
by Ong et al on number of technical grounds. For instance,
area is determined by the site not the panels. Mark points out
that PV efficiencies have increased and will continue to do so and
that the results from Ong et al should have been adjusted
to suit.
Attention turns to the wind farm warming paper. Issue 5 examines the numerical model M+K use to estimate mixing. M+K state that their model matches observed data well. Mark claims the comparisons presented are unable to validate the model.
Issue 6 examines which researchers have or have not undertake particular work in the area. Mark points to his GATOR-GCMOM model in his defense.
Issue 7 concerns the fact that the M+K model is regional not global. They write:
Our paper says nothing about global warming one way or the other. We make no such claims in the paper.
Mark responds that the paper title signals "climate impacts" not "regional climate impacts". Mark then opines that global is implicit in much of the paper. Mark also says two studies that link wind farms to global cooling should have been discussed too.
Issues 8 and 9 continues the theme of local, regional, and global impacts and whether there is a signal in the noise.
Issue 10 turns to the WRF model — which is, according to M+K — an "open community model with public access to the code and documentation". See here (it could well be public domain and unlicensed as the project seems to originate from US agency NCAR):
Mark claims that radiative transfer and water vapor should have been discussed. M+K acknowledge the topic was not traversed but that anyone could read the WRF documentation and determine that these mechanisms are supported. Mark says it is not possible to determine from the documentation how the modeling was run.
In issue 11, M+K point out that GATOR is a private model lacking good documentation. Mark, who wrote 90% of GATOR, says the model is extensively covered in publications, theses, and one textbook (Jacobson 2005) and well established scientifically.
In issue 12, M+K finally say that none of the papers Mark refers
in his earlier critique discuss radiative transfer mechanisms
within GATOR. Mark cites several publications in response.
Closure
I guess that this debate will migrate to academic journals with
time. Until then, I guess the discussion will continue largely on
mailing lists, social media, and science and technology
publications.
But interesting that model openness and documentation are key
points of conflict. Being able to point to a GitHub repo, an open
license, and an associated community would clearly have merit in
this particular context.
with best wishes, Robbie
References
Jacobson, Mark Z (6 October 2018). Response to Reply of Miller and Keith.
Jacobson, Mark Z (5 May 2005). Fundamentals of atmospheric
modeling (2nd ed). Cambridge, United Kingdom: Cambridge
University Press. ISBN 978-0-521-54865-6.
-------- Forwarded Message --------
| Subject: | Re: Recent publications from Miller and Keith, Harvard University |
|---|---|
| Date: | Fri, 26 Oct 2018 07:58:57 -0700 |
| From: | Mark Z. Jacobson <jaco...@stanford.edu> |
| Reply-To: | jaco...@stanford.edu |
| Organisation: | Stanford University |
| To: | Robbie Morrison <robbie....@posteo.de> |
| CC: | Peter Enevoldsen <peteren...@btech.au.dk> |
Hi Robbie,
The Miller and Keith Papers are wrong with respect to their solar and wind installed densities and the impacts of wind turbines on climate, as discussed in detail in this document, which you may not have seen yet
https://web.stanford.edu/group/efmh/jacobson/Articles/I/CombiningRenew/18ResptoReplyMK.pdf
As you can see from the example in this document for the Tule wind farm, their average of 1.52 W/m2 is not even close to correct for the Tule wind farm.
In addition, they calculate around 3 W/m2 (I believe 3.3 more precisely) for the Bull Creek farm. One of the attached figures shows a 3 W/m2 envelope around the actual farm, and clearly there are large amounts of space that have nothing to do with the farm. In fact, one can draw a more precise envelope around clusters of turbines in the farm, and the installed power density is at least 5-8 times larger (although I don't have the exact number, but see second attachment). Miller and Keith's choice to derive their polygon by assuming the large separation between rows counts as spacing area when it is clear additional turbines (or many other landuses, including solar or agriculture or rangeland) can be placed in this spacing area, is a misleading way of calculating spacing area of turbines. In addition, their method results in erroneous areas around the wind farm being calculated. The proper way it to separate a farm into clusters of turbines and account for the reasonable spacing (at least the ground to tip distance if the turbine falls over), as done in the second attachment.
Best regards,
Mark
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/45d7b937-4dd2-ecf3-0f59-dd8546e3de0f%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Robbie Morrison
Hello all (cc: Mark Jacobson and Peter Enevoldsen FYI)
I have it in mind to coordinate a rebuttal on just the German example in Miller and Keith (2018). Scroll down to “consider Germany” to view the original content. A couple of potentially useful references listed below to get the ball rolling. Plus the admittedly dated list of studies on Wikipedia mentioned earlier. I would prefer to work on LaTeX/Overleaf but I supposed Google Docs would have more appeal?
Anyone interested in
signing on? Reply to this thread or email me personally.
cheers, Robbie
References
Ausfelder, Florian, Frank-Detlef Drake, Berit Erlach, Manfred Fischedick, Hans-Martin Henning, Christoph Kost, Wolfram Münch, Karen Pittel, Christian Rehtanz, Jörg Sauer, Katharina Schätzler, Cyril Stephanos, Michael Themann, Eberhard Umbach, Kurt Wagemann, Hermann-Josef Wagner, and Ulrich Wagner (November 2017). Sektorkopplung - Untersuchungen und Überlegungen zur Entwicklung eines integrierten Energiesystems [Sector-coupling: investigations and considerations for the development of an integrated energy system] (in German). München, Germany: acatech, Leopoldina, Akademienunion. ISBN 978-3-9817048-9-1. See Sankey diagram in figure 4 for an overview of system that Miller and Keith investigated.
Miller, Lee M and David W Keith (2018). “Observation-based solar and wind power capacity factors and power densities”. Environmental Research Letters. 13 (10): 104008. ISSN 1748-9326. Open access.
Moraes Jr, Luiz, Christian Bussar, Philipp Stoecker, Kevin Jacqué, Mokhi Chang, and Dirk U Sauer (1 September 2018). “Comparison of long-term wind and photovoltaic power capacity factor datasets with open-license”. Applied Energy. 225: 209–220. ISSN 0306-2619. doi:10.1016/j.apenergy.2018.04.109.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/40ffd65b-ee44-9a94-3c0a-0e95bd564274%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Craig Morris
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/676e7642-79bd-80fa-886a-0a39fb268e83%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Robbie Morrison
Hello all
Reposting with permission. Note also:
- Enevoldsen, Peter and Scott Victor Valentine (1 December 2016). “Do onshore and offshore wind farm development patterns differ?”. Energy for Sustainable Development. 35: 41–51. doi:10.1016/j.esd.2016.10.002.
Robbie
-------- Forwarded Message --------
| Subject: | RE: Antw: [openmod-initiative] Controversial new study on wind energy with findings also misrepresented by Elsevier |
|---|---|
| Date: | Wed, 31 Oct 2018 08:28:30 +0000 |
| From: | Peter Enevoldsen <peteren...@btech.au.dk> |
| To: | Craig Morris <c...@petiteplanete.org>, Robbie Morrison <robbie....@posteo.de> |
| CC: | openmod-i...@googlegroups.com <openmod-i...@googlegroups.com>, Mark Jacobson <jaco...@stanford.edu> |
Good morning Craig,
We have recently conducted an extensive study for Germany where we analysed the technical potential of onshore wind power in Germany. The study was carried out by including all regional distance requirements from buildings, infrastructure, protected areas, etc, which was mapped using new GIS techniques. We furthermore removed areas with certain inclinations and areas where the Global Wind Atlas had less than 5 m/s at 100 meters (Some would argue that we should have used 6). By doing so, it was revealed (as wind turbines can be located on agricultural land and in forests) that Germany has approx. 40,000 km2 (or roughly the size of Denmark) available for new wind project development.
The installed potential is then obviously depending on the spacing density applied😊
|
Peter
Enevoldsen |
BTECH Tel.: 87164700 Mail: bt...@au.dk
|
Robbie Morrison
Hello all
Am repeating my offer that anyone wishing to join me on a
response to Miller and Keith (2018) on the use of power
densities for assessing renewables potentials should contact me
offline. We should be able to make a good case that the article
in question be corrected in print.
The current draft is attached (release 01) and is under my name
alone. This will be the last public release because future work
will continue offline among the authors.
More pointers to the literature would be invaluable. I have yet
to include work by KIT, RWTH Aachen, or open_eGo, for instance.
Can people help out with information, publications, and ideally
related results?
Thanks already to several people who provided suggestions and
references offline.
with best wishes, Robbie
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/6d7d49d3-72ba-f354-e7e6-2e45564b7e62%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Robbie Morrison
Hello all
Two recent articles contribute usefully to the debate on the use of intensive metrics for characterizing wind farms and assessing national wind potentials.
Konetschny et al (2017)
This study (its landing page listed here as FfE 2017), carried out by the Forschungsstelle für Energiewirtschaft (FfE), looks at the near-term wind potential for Germany. The study employs simulated wind farms sited using configuration heuristics to calibrate nonlinear power and energy density functions for later use in a detailed resource assessment. Some interesting points (the translated abstract is in the reference list):
- output power density is Leistungsdichte and MW⋅km−2 = Wm−2
- depending on the location and wind resource, three different turbines are used
- the energy density depends strongly on the size of the wind‑park (as well as the turbine type) due to turbine interactions, see figure 3 (p60)
- if 2% of the land area of each federal state or Bundesland is employed with the optimal turbine type and small wind‑parks favored, then 650 TWh of electricity per year, equivalent to 74 GW continuous, can be generated
- this result is 65% higher than comparable studies, mostly because of the reduced size of the wind‑parks and therefore higher resulting energy densities
The authors identify typical output power densities of 35, 45, and 65 Wem−2, related to the three choices of turbine — two orders of magnitude higher than the 0.5 Wem−2 used by Miller and Keith (2018) for their back‑of‑the‑envelope calculation for Germany.
As noted, Konetschny et al establish their simulated power density metric through ideal turbine choice and configuration and not through historical observation.
On the other hand, it is unlikely that historical US wind farm metrics have much validity when assessing future wind potentials for Germany. Asset vintage, technological development, and attitudes to maintenance (citation needed) mean that historical calibrations must necessarily differ. Land resources are generally more constrained in Europe, providing an incentive for compaction that may not generally exist for sites in the United States. Wind farm sizes also differ, as evidenced by the fact that only one (Fântânele-Cogealac, Romania) is among the world’s 20 largest onshore wind farms, compared with nine in the continental US.
Schwartz (2014)
In a keynote speech, Schwartz (2014) cites an example whereby three wind farm planners overestimated production for a dense wind farm by 20%. This value then provides an indication of the upper bound on uncertainty when planning individual tightly‑spaced wind farms using advanced computer methods.
Discussion
Both results are a far cry from the unsubstantiated 5‑fold or greater descrepency suggested by Miller in his video (Miller and Keith 2018, starting 02:32):
If you are a city, state, or country now implementing plans to becoming 100% renewable, you may need to devote 5 to 20 times more land to wind or solar than your original plans indicated in order to meet your renewable energy targets.
In fact, Konetschny et al head in the opposite direction. While Miller’s comment covers both wind and solar, statements by Miller elsewhere indicate he believes power densities for wind are more errant than for solar. Elsewhere, Miller, as a co‑author, indicates (Gans et al 2012:79):
We conclude that the common method significantly overestimates wind power potentials by an order of magnitude in the limit of high wind power extraction. Ultimately, environmental constraints set the upper limit on wind power potential at larger scales rather than detailed engineering specifications of wind turbine design and placement.
There is concern among meteorologists that an ever increasing number of wind farms will slowly but surely degrade the aggregate wind resource (for instance, Miller et al 2016). On the other hand, that sentiment is somewhat puzzling given that hundreds of wind farms have been designed, commissioned, and monitored over the last 20 years or so and no fall off in wind farm performance nor wind speeds has been reported. Perhaps the effect is only really evident when wind turbines cover the entire land surface, a scenario that meteorologists seem inexplicably drawn toward.
Degradation effects aside, it is hard to reconcile that each camp produces such different estimates for national wind potentials. Miller, for instance, complains that energy modelers fail to account for shading and wake effects when undertaking large‑scale wind assessments. I have not seen recent evidence of this and have yet to see Miller cite a concrete example — with the caveat that I am still working through the literature. Konetschny et al explicitly consider adverse turbine interactions as discussed earlier and err on the side of smaller farms.
Rötzer (2018)
The Miller and Keith paper under discussion has received mainstream press coverage too. Here is the title, tagline, and a sentence from Rötzer (2018) on heise.de (translated):
When wind energy contributes to global warming
A massive expansion of onshore wind farms as part of the move to decarbonization would consume much more area than solar systems because of lower energy density
According to their calculations, if wind farms were to meet US electricity needs and cover one-third of the land area, wind energy would contribute to global warming as the surface temperature would increase by 0.24°C.
The Der Speigel article mentioned earlier and this article were largely negative and riddled with errors of fact and interpretation. Moreover, no countervailing opinion was sought by the journalist concerned in either case. Clearly sloppy work.
Closure
While the example of Germany by Miller and Keith was thin, faulty, and over‑interpreted, other peer‑reviewed material by Miller and colleagues should be examined and compared with current renewables assessments accordingly. That said, I would feel more confident here if Miller’s pronouncements in other contexts were both measured and supported by literature.
Notwithstanding, the wide discrepancies for power density between meteorologists and energy modelers needs studying and resolving. For instance, Miller et al (2016:11169) conclude:
Our results show that the reduction of wind speeds and limited downward fluxes determine the limits in large-scale wind power generation to less than 1 Wm−2.
In contrast, Konetschny et al (2017) employ power density values in the order of 50 Wm−2 as indicated. The calibration of output power density is one issue, including its dependence on turbine packing and wind farm size.
The other issue is resource degradation. It seems that meteorologists think that horizontal momentum in the atmospheric boundary layer — meaning wind for wind turbines — is a somewhat depletable resource. Part of the problem may be that meteorologists have not sufficiently distinguished these two issues: yield and degradation. Nor have they noticed that plastering the planet with wind turbines or other renewable technologies does not produce valid future scenarios nor does it provide information of value to the current debate. Indeed it may well do the opposite, as evidenced by Rötzer (2018).
The WWF study (Matthes et al 2018) cites inverse power densities in m2W−1. I haven’t had a chance to look at the values though.
Finally, if modeling teams on this list could calculate and post their output power densities, howsoever aggregated, that would be great! Also, any reported observed values would also be welcome.
The wind technology lexicon listed below provides useful definitions.
Work on a more considered response to Miller and Keith (2018) continues offline.
All translations provided here are unofficial.
with best wishes, Robbie
References
FfE (6 June 2017). Potenzielle Leistungsdichte und Stromerzeugung von Windparks Anteil der regionalen Windstromerzeugung am Verbrauch für ein “2 %-Szenario” [Potential power density and power generation of wind farms share of regional wind power generation in consumption for a “2% scenario”] (in German). Forschungsstelle für Energiewirtschaft (FfE). Munich, Germany.
Gans, Fabian, Lee M Miller, and Axel Kleidon (2012). “The problem of the second wind turbine: a note on a common but flawed wind power estimation method”. Earth System Dynamics. 3: 79–86. doi:10.5194/esd-3-79-2012.
Konetschny, Claudia, Tobias Schmid, and Fabian Jetter (May 2017). “Potenzielle Leistungsdichte und Stromerzeugung von Windparks Anteil der regionalen Windstromerzeugung am Verbrauch für ein “2%-Szenario”” [“Potential power density and power generation of wind farms share of regional wind power generation in consumption for a “2% scenario””] (in German). Energiewirtschaftliche Tagesfragen. 67 (5): 59–62. Scanned document. Abstract (unofficial translation):
Targets for wind energy utilization are given in different metrics [literally: units of measurement]. In addition to the number of plants, the power and the annual power generation, the indication of an area to be designated for wind power use is increasingly employed. However, the relationships between area consumption, installed capacity and amount of electricity generated are not nearly linear. Based on fine-grained modeling of the installation of wind turbines with the aid of the Wind Scenario Tool WiSTl developed by the Forschungsstelle für Energiewirtschaft (FfE) [Research Center for Energy Economics], the article provides recommendations for the conversion of the designated area. In addition, based on a 2% scenario, the share of power generation from wind turbines is determined relative to the electricity consumption of each German federal state.
Matthes, Felix Chr, Franziska Flachsbarth, Charlotte Loreck, Hauke Hermann, Hanno Falkenberg, and Vanessa Cook (October 2018). Zukunft Stromsystem II: Regionalisierung der erneuerbaren Stromerzeugung : Vom Ziel her denken — Version 1.1 [Future electricity system II: regionalization of renewable electricity generation: thinking on the goal — Version 1.1] (in German). Berlin, Germany: WWF Deutschland. ISBN 978-3-946211-22-8. Make sure you locate this version.
Miller, Lee M and David W Keith (2018). Observation-based solar and wind power capacity factors and power densities. Cambridge, Massachusetts, USA: Harvard University Center for the Environment. Video 03:48.
Miller, Lee M, Nathaniel A Brunsell, David B Mechem, Fabian Gans, Andrew J Monaghan, Robert Vautard, David W Keith, and Axel Kleidon (8 September 2015). “Two methods for estimating limits to large-scale wind power generation”. Proceedings of the National Academy of Sciences (PNAS). 112 (36): 11169–11174. ISSN 0027-8424, 1091-6490. doi:10.1073/pnas.1408251112.
Rötzer, Florian (10 October 2018). Wenn Windenergie zur Klimaerwärmung beiträgt [When wind energy contributes to global warming] (in German). Telepolis. Hannover, Germany.
Schwartz, Herbert (16 September 2014). Ertragsverluste durch Windparksituationen — Impulsvortrag [Yield losses due to wind farm situations — Keynote speech] (in German). Berlin, Germany: Bundesverband Windenergie (BWE). Workshop of the Windgutachterbeirats [Wind Assessment Advisory Council] of the [Bundesverband Windenergie] German Wind Energy Association.
wind-lexikon.de (ongoing). Das Lexikon zur Windtechnologie [Wind technology lexicon] (in German). wind-lexikon.de. Germany. Good source of definitions.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/3920ba00-0e03-7701-22da-26dc4f977326%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Robbie Morrison
Oops, correction:
- the
energypower density depends strongly on the size of the wind‑park (as well as the turbine type) due to turbine interactions, see figure 3 (p60)
Actually both depend thus, Robbie
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/96a403be-c318-8074-4b23-86cf737d64ee%40posteo.de.
For more options, visit https://groups.google.com/d/optout.
Ken Caldeira
Fuel
Mtoe
oil
119.8
natural gas
<td style="border-top:none;border-left:none;border-bottom:solid #CCC
Robbie Morrison
Hello Ken, all
Both interesting and
serious. Quoting from Miller et al (2015:11169) (which I
dated as 2016 in my earlier post):
Wind turbines generate electricity by removing kinetic energy from the atmosphere. We show that the limited replenishment of kinetic energy from aloft limits wind power generation rates at scales sufficiently large that horizontal fluxes of kinetic energy can be ignored. We evaluate these factors with regional atmospheric model simulations and find that generation limits can be estimated from the “preturbine” climatology by comparatively simple means, working best when the atmosphere between the surface and hub height is naturally well-mixed during the day. Our results show that the reduction of wind speeds and limited downward fluxes determine the limits in large-scale wind power generation to less than 1 W⋅m−2.
Konetschny et al (2017) were looking at 2% area, spread throughout Germany because of their Bundesland constraint. I take it that the 1 W⋅m−2 is spatially averaged .. so 1.0/0.02 = 50 W⋅m−2. Okay, that's the back‑of‑the‑envelope stuff I've been complaining about!
Clearly an unresolved
issue there. We did have one person from the Meteorology
Department, University of Reading, United Kingdom attend one of
our workshops. Later I exchanged email with David Brayshaw.
Perhaps I should see if those folk can contribute anything to
this question? Do you know whether other researchers are working
on the topic?
with best wishes,
Robbie
--
You received this message because you are subscribed to the Google Groups "openmod initiative" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openmod-initiat...@googlegroups.com.
To post to this group, send email to openmod-i...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/openmod-initiative/CAKNUXC0_bad408j%3D%3D7_pTO9pU0-4yNEznndY8eVN5BFCN_qOVA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.