---------- Forwarded message --------- From: Wake Smith <w...@crowsven.com> Date: Sun, 6 Jan 2019, 21:23 Subject: RE: [geo] Stratospheric aerosol injection tactics and costs in the first 15 years of deployment - IOPscience To: Andrew Lockley <andrew.lock...@gmail.com> Cc: geoengineering <geoengineering@googlegroups.com>
Dear Andrew (& group), Firstly, thank you for the thorough and thoughtful questions. I am happy to dig in further on these details with knowledgeable correspondents. Secondly, I should note that I do NOT consider this paper to have been the final and definitive word on early deployment tactics, but rather simply (and hopefully) a forward step from the essential work done earlier by McClellan et al and others. “McClellan” (as I will hereinafter refer to their paper) remains foundational and I started my explorations by meeting with both McClellan and Keith and picking up the ball where they laid it. I am comfortable with our paper insofar as it went, but I acknowledge there to be many yet still unanswered questions which I intend to address in subsequent undertakings. Thirdly, I am speaking here only for myself, though Gernot will undoubtedly weigh in additionally as he sees fit. Starting with your “Tilmes +5k” question, I should note that our paper diverged from McClellan at the outset by choosing a specific mission and then considering platforms to fulfill that mission and only that mission. McClellan on the other hand considered deployment altitudes ranging from 18 – 30 kms, targeting the lower part (18 – 25kms) of that range. Table 2 surveys an even wider range, from as low as 40kft (~12.2 km) up to 100 kft (~30 km). So wide a spectrum of possible injection altitudes naturally leads to a wide variety of platforms suitable to address at least some of the part of that spectrum and contributed to an impression that there were many ways to “skin the cat” as it were. More specifically, this implied that some sub-stratospheric altitudes were nonetheless acceptable for deployment even though the text of the paper called for injection above the tropopause. From the standpoint of the grubby aviation guys simply trying to fly the mission, altitude is the critical parameter here, so more specificity was required in order to zero in on a platform choice. We therefore chose to define a much more specific mission that always deployed well into the stratosphere, which in turn led us to a more specific platform recommendation. The mission we chose was injections as high as 65k ft (~20km), and we sourced this mission requirement from MacMartin/Tilmes/Kravitz. To be clear, this does not mean that all injections would necessarily achieve this altitude – one might choose lower on particular days and at higher latitudes – but the maximum injection altitude anywhere defines the altitude threshold for the platform design. So, why didn’t we consider the engine alternative you note? Because it was not necessary to achieve the defined mission. The above of course begs the question as to whether we chose the right mission, or whether we should have instead chosen various alternative missions, such as the “+5k” alternative. As regards the mission we chose, we had to start somewhere, and this seemed (and still does seem) like the right place to initially plant the stake, in part because it was based upon such a well-respected prior paper. That said, one of my personal projects for 2019 is to drill down further on the question of necessary deployment altitude to further clarify the dynamics that define both the release altitude and the migration of the material after release, so more to come on this. However, the primary question I am pursuing there is whether we can in fact deploy LOWER, not higher. I am mindful of the ~50% further radiative benefit that Simone notes would be achieved by an additional 5k of altitude, but that seems somewhat marginal on top of the ~50X benefit we achieve by ensuring we are in the stratosphere rather than the troposphere, and that 5k would come at a very high cost. You imply that one could get there by simply strapping different powerplants on the wings, but I highly doubt that. 20 kms is the ragged edge of what can be achieved with traditional fixed wing, self-propelled aircraft – none of the Global Hawk/U2/SR-71/WB-57 get materially higher than that, and each of those with a mere ~1 ton payload. Getting 25 tons up to 20 kms would be unprecedented, though achievable, I believe. Getting up to 25kms WITH A HEAVY PAYLOAD (please don’t send me artwork of the Perlan and such) means we are no longer flying aircraft, but moving to other platform types that we have demonstrated come with a minimum 10X cost differential. 10X more cost to get 0.5X more benefit doesn’t seem like a sensible trade – BUT, if someone knows something I don’t, please advise (I will stop repeating this, but this sentiment is general to the dialogue here). I don’t share your understanding that mid-air refueling would be cheaper option – quite the opposite. Mid-air refueling is done to supplement range at a materially added cost, not to save cost. SAIL would be a relatively short range aircraft not for the same reason as fighter jets (to reduce weight so as to enhance speed/maneuverability), but because it needs little range. All it need do is pop up to altitude, deploy more or less above the base, and descend with some reserve fuel. I see no benefit to the further complication (and flight time) required for mid-air refueling. Regarding Delft and cost, the first thing to note is that our paper clarifies that the vast majority of the costs in the first 15 years are operating costs, not developmental costs, and we contribute a vastly more detailed and reliable estimation of those operating costs, so your question goes to the much lesser cost category. Nonetheless, since those are the FIRST costs, they are particularly important in defining the financial threshold necessary to get started. So – the key drivers of reduced developmental cost are engines as you note, and then certification cost. Both McClellan and Delft (which itself relied heavily on McClellan) used comparisons to prior development programs of commercial airliners, such as (in Delft) the B787 and A380. Commercial airliners require enormous amounts of optimization in order to reduce weight/drag/fuel consumption and to enhance reliability – all to beat the operating cost of the predecessor aircraft they are replacing by 20% and the competitor’s aircraft by a few percent. They also need to explore every corner of the flight envelope since they are expected to be operated all over the world by scores of operators in nearly every conceivable mission profile. All of that adds dramatically to developmental cost, and none of that applies here. We are assuming that SAIL commences life as the global sole-source platform for SAI (so no need to out-do the competition) and that it will be designed for this mission only (so not a “flexible” platform that would likely be repurposed to other uses if the world abandons SAI). In these regards, this resembles a military development program much more than a commercial one. We also assumed that the aircraft could operate either under an experimental or a military certification. Confirming this latter assumption (ie, the likely “cert basis”) is another vector of further personal research, as is the “governance” assumption implied in the “global sole-source” comment. Our cost-per-deployed-ton is as you note nearly identical to McClellan, but is arrived at via very different methodology. You ask how our aircraft design is different, but McClellan doesn’t posit a design – it analyzes a set of design parameters to drive out an approximate size (“similar to a G200”) and costing for aircraft targeting altitudes from 40kft to 100kft. Our aircraft is ~6X the weight of a G200, but it is different chiefly in that we lay out a much more specific set of dimensions, weights, and capabilities. Regarding manned vs unmanned cockpit, if this bird were actually to enter into service in 2033, my guess is that it would be unmanned. However, in the current and immediately foreseeable future, it would be much cheaper to certify it as a manned vehicle. The cert program and requirements for large unmanned aircraft are still evolving and will be quite different a decade hence in ways that are hard to predict. Rather than try to gaze into that crystal ball and guess how such a cert program might be administered and therefore what it might cost, we stuck to the more predictable cert path for manned vehicles. As for your “highly encapsulated” single purpose airport concept, the direct answer is – no, we did not analyze that and I don’t see a need to do so. That would add quite substantially to the cost (after all, we anticipate four and soon eight bases) for little benefit. These are not particularly dangerous aircraft, either for the operators or the people on the ground. If there were a need to segregate these aircraft from other traffic, my guess is that it would arise from concerns about security rather than safety. Sulfuric acid is ~4X the mass of molten sulfur, so not only would it multiply the cost by roughly 4X, but it would reintroduce safety concerns that molten sulfur ameliorates. I claim no expertise on particle size considerations, but until someone clarifies that hauling sulfuric acid is worth the costs and risks, I don’t see the case for that. On engine mods, Rolls provided a considered but preliminary estimate that it would require $210 - $420MM to qualify their BR710/725 model for this operation, and I chose a value in the high end of that range ($350MM) to provide some margin. I cannot provide a breakdown, but can clarify that they did not believe any changes to the core would be required. Rather, the main mods would be to the fuel storage and flow system. Larger yet from a budgetary perspective would be testing and recertification costs. There is just one test cell in the world that can recreate the conditions experienced at such high altitudes as this, and it is extremely expensive to rent. Rolls anticipated an extensive program there with three test engines. GE and Pratt have existing engines that already operate at roughly this altitude, so the costs would likely be much lower (perhaps negligible) in the case of their engines. SpaceX has been unresponsive to my inquiries (I had better luck elsewhere), so I have had to rely upon public data regarding their costs. However, the costs for all the rocket alternatives are so far (like 50X) beyond the cost for aircraft that further banging on those doors does not seem a productive use of time. Airships were also easy to cross off the list based upon technological immaturity, and electrical launch, even more so. These and many other lofting solutions may well prove better and cheaper than the SAIL aircraft in the long run, but we intentionally sidestepped the sinkhole of trying to guess what whizz-bang solutions will appear several decades from now by limiting our focus to the initial deployment efforts in the not-too-distant future. Guns is the one competing technology which I have difficulty crossing off the list. You may note that this is the one section where we simply fell back upon McClellan’s work because we were unable to obtain data that provided substantially improved insights – but not for lack of trying. Rather oddly, big guns are a technologically sleepy (bordering on dead) field. The enormous battleship and rail guns of the mid-20th century have all been retired and replaced on succeeding generations of ships by vastly smaller guns, mostly because missiles have replaced them. The US Navy (by far the world’s biggest big-gun customer) is down to just one supplier (BAE Systems) who has little in development that might suit the needs of an SAI program (and even their most promising program was defunded almost a decade ago). Both the Navy and BAE Systems declined to provide information on what are inactive but still classified programs or to speculate on what could potentially be developed, and McClellan’s unfavorable cost comparisons justified bypassing this alternative. And yet….in this case, it’s clear that even the WWII-era technology would haul the required payloads to the necessary altitudes, and eliminating hypothetical 21st century guns based on 75 year old cost comparisons seems unreliable. Neither do I consider this to be in the “technologically immature” category – the products don’t now exist, but there is no big, unforeseeable breakthrough required to produce them – it’s simply engineering. Parenthetically, I have very little faith in the “reusable shell” idea, and in the case of guns one WOULD have to haul the final product, so that if that is SO2 we are back 2X mass/cost problem at best. My hunch is that with more complete data, we would more confidently eliminate guns, so I am not greatly worried about this gap in our data, but it’s a gap nonetheless. Efforts by others to fill it would be most welcome. More to follow in 2019 on several related topics, and I should reiterate for the record that we hope the world will find a way to deviate from a path that would require SAI deployment in the first place. Nonetheless, since no signs of such a deviation are yet evident, reducing uncertainties in respect of Plan(s) B seems very much in order. Wake Smith w...@crowsven.com 914 649 7722 *From:* Andrew Lockley [mailto:andrew.lock...@gmail.com] *Sent:* Tuesday, December 25, 2018 7:33 PM *Cc:* geoengineering <geoengineering@googlegroups.com> *Subject:* Re: [geo] Stratospheric aerosol injection tactics and costs in the first 15 years of deployment - IOPscience Wake / Gernot I have some questions on your paper. I thought it would be best to pose them in public, in the hope that others will be able to read any reply. I apologise in advance if there are elements of your paper I have not fully understood. Citing Tilmes, you suggest a + 5 k altitude change would be beneficial , but suggest engines are a limiting factor . The BAE Systems Sabre engine is designed for high-altitude use. (David Keith was previously critical when I suggested the use of this system). If you have considered this engine type (or similar), why did you disregard it? Mid-air refueling is an established technology. Your paper does not discuss the idea of conveying fuel or payload in flight. The high-altitude aircraft you propose would be less fuel efficient and more expensive than conventional tankers. These factors imply that any element of the process that can be outsourced to tankers would represent a significant cost saving. Was there a reason why you did not consider a two stage approach? The TU Delft aircraft appears superficially similar to yours in design, save the use of custom engines. Why does your Design come out at such a dramatically lower cost? Your proposed costings are almost identical to the new aircraft design proposed by Mcclellan. Your paper does not give much detail on why your Design would be different , and what advantages it would have. Could you please elaborate on this? You plan a manned aircraft, but the reduced safety concerns of a drone mean that certification costs are potentially lower, particularly if the planes were flown from isolated, single purpose airports (where any crash would be unlikely to cause damage or injury). Did you analyse any such highly encapsulated operational model? On board conversion of sulphur is unlikely to allow very fine control over droplet size. Any outsize aerosols are both much heavier, and much less effective, than an ideal monodisperse spray. Did you consider the alternative of carrying sulphuric acid, so you could inject monodisperse aerosols? If so, what are the cost implications? You give very little detail on the engine modifications necessary. Could you please elaborate on their nature, and offer a breakdown of anticipated cost? As regards alternative technologies, I have some further questions: Your consideration of costs from SpaceX appears to be based on their space launch technology. This is inherently a low volume operation. SpaceX have also proposed a suborbital passenger service, which would have far higher flight volumes - and thus far lower marginal costs per launch. This appears not to have been included in your analysis. Have you done any side calculations , based on realistic cost assumptions for adapting SpaceX suborbital passenger rockets ? You suggest that airships are at too early a stage of development to be practical. Hybrid Air Systems already have a flying vehicle of the type required, albeit one not adapted to this specific job . Did you examine this firms technology? If so, what were your reasons for rejecting it? Maclellan's paper considered gun launch, but did not consider obvious opportunities for costs savings. These include: reusable shells; and converting the guns from specialised solid propellant bags, to natural gas / hydrogen fuel, with air as an oxidizer. Further , guns allow much higher altitudes than aircraft, which you advised is desirable for reasons of efficiency. Such modifications would imply a cost reduction of approaching one order. Is there a reason you have elected not to optimise gun designs, in your analysis? Finally, you make no reference to any electrical launch technology. A cursory look at hyperloop suggests that it can be modified to attain approximately the launch velocities required. Did you consider this, or similar electrical launch? If so, why did you reject it? I look forward to receiving any response you are able to send. Andrew On Sat, 24 Nov 2018, 14:35 Douglas MacMartin <dgm...@cornell.edu wrote: For context, the “huge expense” you refer to below, for the first 15 years of deployment, is about 1.5x the estimated cost of the Camp fire in California last week. Or, 15 years of deployment (including development costs), are about 15% of the costs in the US alone from the 2017 hurricane season. And certainly far cheaper than actually solving the problem by pulling out the CO2. Lots of reasons to be concerned about SAI, but as far as costs are concerned, the appropriate concern should be that it is too cheap, and that cost won’t present enough of a barrier to deployment. (And as I’ve pointed out before, saying this doesn’t “solve” the climate problem is like pointing out that air bags don’t “solve” the problem of having car accidents, or a million other analogies. Of course it doesn’t. No-one says it does. But it could reduce impacts and prevent lots of climate damages. Until we are certain that the climate problem can be “solved” by other means, it would be premature to dismiss something that has the potential to limit damages.) *From:* geoengineering@googlegroups.com [mailto: geoengineering@googlegroups.com] *On Behalf Of *Franz Dietrich Oeste *Sent:* Saturday, November 24, 2018 6:49 AM *To:* andrew.lock...@gmail.com; geoengineering@googlegroups.com *Subject:* Re: [geo] Stratospheric aerosol injection tactics and costs in the first 15 years of deployment - IOPscience Thanks to Wake Smith and Gernot Wagner for their work! Their paper may open our eyes to the probable unsuitability of the climate influencing tool Stratospheric Solar Radiation Management (SRM) or as named by the authors Stratospheric Aerosol Injection (SAI): SRM shall act within the stratosphere 20 km above the ground. To gain a temperature reduction of 0.30 K in 2047 it needs a yearly uplift to this height of 1,5 million tons of sulfur. The sulfur shall be burned by new kind of aircrafts in situ to gain gaseous SO2 (boiling point -10 °C) which becomes transformed by oxidiation and hydration to about 6 million tons aerosol made of a rather concentrated sulfuric acid - per year. This aerosol shall spread around the globe and mirror parts of the sun radiation back into the space. With the existing aircraft design sulfur lifting to these heights is impossible. A new kind of aircraft needs to be developed to do the job. This new aircraft should be able to lift a payload of 25 tons of liquid sulfur to 20 km above the ground then keeping at this height and burn there the sulfur load which emits with the flue gas as SO2. About 60 000 flights per year are necessary to gain the global temperature reduction of 0,30 K. Thankfully this article discusses very clearly within chapter 6 that such activities could not remain undetected. Their conclusion is that it would be rather impossible that those activities remain undetected or might kept as a secret. According to this low result of 0,30 K global temperature decrease gained by this huge expense and 1,5 Million tons of sulfur burned in the stratosphere the SRM method seems completely unsuitable to solve our climate problem. Not only that the SRM method does not reduce any of the increasing levels of the essential greenhouse gases CO2 and methane, it surely increases the CO2 gas level. Any reduction of the sun radiation at the surface decrease the assimilation by which plants transform CO2 into organic C and oxygen. Further SRM would increase the methane level by decreasing the UV radiation dependent hydroxyl radical level which acts as a degradation tool to methane and further volatile organics because the sun radiation decrease by SRM concerns particularly the UV fraction. It is my very hope that this article helps to reduce the hype about SRM. Franz D. Oeste ------ Originalnachricht ------ Von: "Andrew Lockley" <andrew.lock...@gmail.com> An: geoengineering@googlegroups.com Gesendet: 23.11.2018 16:36:27 Betreff: [geo] Stratospheric aerosol injection tactics and costs in the first 15 years of deployment - IOPscience *http://iopscience.iop.org/article/10.1088/1748-9326/aae98d/meta* <http://iopscience.iop.org/article/10.1088/1748-9326/aae98d/meta> Stratospheric aerosol injection tactics and costs in the first 15 years of deployment Wake Smith1 and Gernot Wagner2 Published 23 November 2018 • © 2018 The Author(s). Published by IOP Publishing Ltd Environmental Research Letters, Volume 13, Number 12 Download Article PDF DownloadArticle ePub Article has an altmetric score of 157 Abstract We review the capabilities and costs of various lofting methods intended to deliver sulfates into the lower stratosphere. We lay out a future solar geoengineering deployment scenario of halving the increase in anthropogenic radiative forcing beginning 15 years hence, by deploying material to altitudes as high as ~20 km. After surveying an exhaustive list of potential deployment techniques, we settle upon an aircraft-based delivery system. Unlike the one prior comprehensive study on the topic (McClellan et al 2012 Environ. Res. Lett. 7 034019), we conclude that no existing aircraft design—even with extensive modifications—can reasonably fulfill this mission. However, we also conclude that developing a new, purpose-built high-altitude tanker with substantial payload capabilities would neither be technologically difficult nor prohibitively expensive. We calculate early-year costs of ~$1500 ton−1 of material deployed, resulting in average costs of ~$2.25 billion yr−1 over the first 15 years of deployment. We further calculate the number of flights at ~4000 in year one, linearly increasing by ~4000 yr−1. We conclude by arguing that, while cheap, such an aircraft-based program would unlikely be a secret, given the need for thousands of flights annually by airliner-sized aircraft operating from an international array of bases. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com. To post to this group, send email to geoengineering@googlegroups.com. Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com. To post to this group, send email to geoengineering@googlegroups.com. Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com. To post to this group, send email to geoengineering@googlegroups.com. Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout.
