Categories
Open Thread

Open Thread 26

Time for a new open thread, since apparently the previous one is now loading a little slowly… I’ll close the old one to comments, so please continue discussion here.

As for the quiescence of BNC over the past few months, well, I’ve been travelling — what can I say? But I have a new post to put up tomorrow, and a few others in train.

The Open Thread is a general discussion forum, where you can talk about whatever you like — there is nothing ‘off topic’ here — within reason. So get up on your soap box! The standard commenting rules of courtesy apply, and at the very least your chat should relate to the general content of this blog.

The sort of things that belong on this thread include general enquiries, soapbox philosophy, meandering trains of argument that move dynamically from one point of contention to another, and so on — as long as the comments adhere to the broad BNC themes of sustainable energy, climate change mitigation and policy, energy security, climate impacts, etc.

By Barry Brook

Barry Brook is an ARC Laureate Fellow and Chair of Environmental Sustainability at the University of Tasmania. He researches global change, ecology and energy.

784 replies on “Open Thread 26”

US federal government subsidies:
solar = $280/MWh
Wind = $30/MWh
Nuclear = $2.10/MWh

Click to access subsidy.pdf

Solar is subsidies over 100x more than nuclear and wind more than 15x.

And this is only part of the subsidies and other market distortions that favour wind and solar

Like

In the previous thread is a link to an Australian article. Here is a quotation therefrom:

“Our latest research, which corroborates previous work, shows the technology already exists to solve many of the remaining questions around technological capability. For instance, the fact that wind and solar don’t generate electricity when the wind isn’t blowing and the sun isn’t shining can be dealt with by installing a network of diverse generators across a wide area, or by increasing our use of energy storage.”

Please note the word “or” in the last sentence of the quotation.

Until a few years ago, here in the U.S. it was also asserted, also without proof, that a diverse network of renewables over an area would guarantee adequate reliable power. So far as I can see, that assertion has been abandoned here in the U.S.; it is now widely assumed that energy storage will be able to make renewables reliable even though an adequate energy storage technology has not been demonstrated except in the limited areas where pumped storage is practical.

The Australian article merely ASSERTS that research indicates that diversity over a wide area will result in reliable power. It does not state how that research was done. The ONLY proof which I would accept, either here or in Oz, would be to have wind and insolation sensors in many of the places where actual installations would be reasonable, and carefully analyze the data over a period of AT LEAST one year to determine if the resulting power would ever be insufficient. Perhaps that has been done but if so, I am not aware of it.

To me it seems that the assumed practicality of renewables (except for hydro and possibly geothermal) is more a matter of faith than science. Considering the strong commitment of Deutschland and a few other European countries to renewables, if renewables alone could actually be made to produce reliable power, some country by now would have succeeded in doing so. Yet, I know of no country which has succeeded in that endeavor. Occasional periods of a few hours here and there where renewables have provided adequate power do not count!

Although I do have reservations about nuclear power, especially our current nuclear technology, it appears that only nuclear power can make it possible to reduce CO2 emissions to an acceptable level and still provide adequate power.

Like

There is a subject that seems to be taboo in mainstream media or most serious blogs, yet is really starting to bother me now.
It is this:
What do ‘we’ do, now that it’s quite clear that neither of Australia’s major political parties are going to take appropriate and necessary action on climate change in an acceptable timeframe?

And who is ‘we’ at this point in Australia’s history?

Liked by 2 people

@Chris Sanderson:
IMHO at least the top 4 parties in Australia are blind to the actions necessary to deal with climate change, possibly because climate change, its effects and the actions which might address the problems have become so divisive that there is a perception that any policy will lose more votes than it will gain, hence no policy is the best policy, especially once key emotional triggers such as renewable, nuclear, solar and wind are mentioned.

A few years back, Barry Brook co-authored “Why Vs Why – Nuclear Power”, which attempted to initiate a public discussion about matters nuclear, not based on emotion and preconceptions but on rational consideration of the facts. It was neither a failure nor a success, in part because the other co-author, Ian Lowe, challenged the foundations of Barry’s arguments. They didn’t agree enough about the facts. So the fire of productive discourse remained unlit.

There are plenty of blogs out there with political stances. BNC’s focus has been on the “How and why”, discussing as objectively as possible the objective truths. The science, if you will. With references to back up claimed facts.

This is as it should be, but we are not all engineers, scientists, statisticians and economists.

As John Cook of Queensland Uni and his MOOC “Making sense of climate science denial” course have concluded, the issue is not science but the psychology of denial.

https://www.edx.org/course/making-sense-climate-science-denial-uqx-denial101x-1#.VIDuaaMcSCr.

I don’t have the answer that Chris seeks, but current politicians follow public opinion, they don’t lead it.

So I rephrase the question to “Why isn’t the public demanding that the government giving first priority to addressing the challenges of climate change?”

Maybe it isn’t a party issue, is a public perception issue. More psychology.

Like

Peter Lang — From which tables in the subsidy report did you obtain the data which led to the astoundingly high subsidy for solar? Thank you in advance.

Like

David Benson, I assumed Peter derived the $ subsidy / MWh figure from tables ES2 (Quantified energy-specific subsidies and support by type, FY 2010 and FY 2013) and ES5 (Measures of electricity production and growth)

From this, I calculated solar at $280/MWh and nuclear at $2.10/MWh but I got a slightly higher figure for wind – $35/MWh

I tried using table ES4 (Fiscal Year 2013 electricity production subsidies and support) but only nuclear corresponded with Peter’s figures as the solar subsidy amount is different because some of the subsidies are used for non-electrical (eg transport and direct heat) applications.

If ES4 is used (which makes more sense, given the units), the $ subsidy / MWh figures for solar falls to $231 (nuclear and wind remain at $2.10 and 35, respectively)

Like

Greg Kaan — Thank you. I will revise accordingly. A critic states that the solar subsidy, forgoing some capital costs, ought to be divided by the lifetime of the panels, nominally 20 years. Do you agree?

Like

Greg Kaan,

Yes, the figure for wind is $35/MWh, not $30/MWh. it was a typo. i noticed it as soon as I posted it, but didn’t want to making another comment with a correction. I am encouraged, that at least one person following the thread took the trouble to check the calculations, as I did when this chart was first posted by an economist on another blog site. Thank’s Greg. :)

Liked by 1 person

Peter Lang & Greg Kaan — Having an opportunity to study the Summary and some of the footnotes I conclude that to fairly compare the subsidies one should divide three quarters of the subsidies for wind and solar by 20 in order to obtain a direct comparison to the nuclear subsidy. This exercise places the wind subsidy at around $10/MWh amortized over the nominal 20 year life of the turbines. For solar about $66.41/MWh, very high as that figure is typical wholesale spot price.

Like

Greg Kaan — From a footnote in the summary. The subsidies in effect pay for capital costs of units built in that year; rather about 3/4 of the subsidies do. So amortizing over the lifetime for that 3/4 is suggested by the footnote.

Whatever, the subsidies for solar are disproportionate.

Like

I look forward to the costs of Watts Bar (and Vogtle and Summer in the future) marking down the nuclear subsidy figure. A 50 year lifespan will dilute the capital cost proportion quite nicely

Like

Terrestrial Energy, a Canadian startup, seems to be planning to build one of their 190 MWe IMSR units on the Idaho National Laboratory reservation. No indication who the lead customer is or how they will ever obtain an NRC license. From their website the concept appears quite good.

Like

Hi, Greg. A link within your referenced article leads to the best blunt discussion of the future of the NRC that I have seen. If NRC doesn’t find a way to become effective, they are going to face increasing competition from the likes of China. This is already dragging the centre of the entire nuclear power industry, especially for SMR’s and thorium fuels, away from USA. This could be forever.

A growing number of engineers interested in a future in the SMR industry will need to learn mandarin, which is no simple task, and to work for Chinese masters, which from my direct professional experience is “interesting”. My daughter gave it her best shot, including a year attending school in Taiwan followed by two years study of the language in an Australian at university. She couldn’t master it.

https://clearpath.org/jays-take/nuclear-innovation-isnt-welcome/

The other alternatives, Russia and Korea amongst them, are no more attractive from a Western point of view.

I really hope that the Americans stay in the driver’s seat, but can they?

Like

A Professor of engineering at ANU is asserting:-

It turns out that we can maintain grid stability in Australia with a scenario of {90% wind+PV, 10% existing hydro+bio+other} by adding in off-river (closed loop) PHES plus a high voltage DC cable down the east coast to transfer wind and solar between north and south (which tend to have different weather systems). The LCOE we calculate is <$100/MWh – which is less than fossil or nuclear or any other renewables.Importantly, wind, PV, PHES and HVDC are in >100GW deployment so no heroic assumptions are required (unlike with most other low emission technologies).PV and wind cost $65-85/MWh in Australia as revealed by the price discovery programs run through the ARENA Large Scale Solar and the ACT reverse auction programs.Moving rapidly to 100% renewable electricity will not be difficult or expensive. Then follows electrification of cars and urban heat.

https://theconversation.com/despite-the-funding-cut-arenas-glass-is-still-half-full-heres-how-to-spend-the-money-65351#comment_1080424

Psychologically, I want to believe him, but you can see how I replied.

Like

Prof Blakers is a specialist in the field of solar cell development where he is eminently qualified, making him an electronics engineer rather than electrical. I have severe doubts as to his expertise in analysing grid stability, especially since his involvement in the Beyond Zero Emissions group would indicate an ideological bent toward renewables. I would say his opinion is worth more than Mark Diesendorf’s but that not saying too much – again IMO.

The argument he presents is basically the one that Mark Jacobson of Stanford University and The Solutions Project puts forward for the USA – conceptually appealing but try talking to the people who get their hands dirty about the reality.

wind, PV, PHES and HVDC are in >100GW deployment so no heroic assumptions are required

That says it all, really. Nothing heroic about deploying > 100 GW of wind, PV, PHES and HVDC, indeed!! Total generation capacity current deployed in Australia would be of that order, with the vast majority being coal and gas power stations.

Like

Greg, Prof Blakers seems to be wide of the mark with his 100GW figure. The current NEM generator capacity is very roughly 35GW fossil fuels and 10GW hydro, solar and wind.

We will have to wait for the paper to be published in order to see the assumptions, model configuration and input data.

I have absolutely no idea where Prof Blakers is deriving his assumption that the southern states will be supplied by renewables in Qld when the weather is poor in the southern states – as is regularly the case for a few days at a time.

Currently, Queensland has a fair penetration of rooftop solar but very little other renewables.

Is one assumption that 50GW of new renewables will suddenly appear north of the Tweed River and be conected to the lower states by multiple HVDC circuits?

If so, that is truly an heroic assumption.

My 45 GW NEM figure came from here:
https://www.aemo.com.au/Electricity/National-Electricity-Market-NEM. Breakdowns of this state by state and by generation technology are available from that site; look for AEMC/Generation/Map.

Current Qld installed capacity is about 12GW fossil fuelled plus 2GW rooftop PV.

Like

From Robert Hinds on the Andrew Blakers posts.
I mentioned Professor Andrew Blakers in a multi subject post on the end of open thread 25. This has been picked up on thread 26 and led to several responses. I heard his interview on ABC breakfast radio and decided to email him. He replied promptly to two emails and I have included them on this post unedited.

Dear Sir. This email is in response to an interview you did on 891 Adelaide breakfast radio recently. It appears from your ANU website that you specialize in renewable power and its storage systems. Regarding efforts being made to reduce greenhouse gases from electricity production, the alternatives appear to be nuclear or renewables. (When access to mega hydro is ruled out.) The Achilles heel of nuclear is permanent disposable of used fuel. The Achilles heel of renewables is intermittency. Could you guide me towards research that would make renewables storage viable. The need for Gigawatt hours of storage for consecutive days is not even being approached by current systems. What innovative systems are in the pipeline over the next few decades that might meet these requirements? Regards.
REPLY. We have preliminary results from our integration studies, looking at the Australian national electricity market, and a scenario with 90% wind+PV, 10% everything else (existing hydro and biomass). We take 30min data for demand, wind and sun and find that we can maintain grid stability by adding in off-river PHES plus a high voltage DC cable down the east coast to transfer wind and solar between north and south (which tend to have different weather systems). The LCOE we calculate is <100/MWh, compared with wind (US$64/MWh) and PV (US$84/MWh). The wind and PV costs are derived from price discovery programs from Governments. In other words, achieving grid stability adds about one third on top of the cost of energy generation. We will publish this work in the next few months.
South Australia has excellent pumped hydro opportunities – the steep hills east of the gulf outside nature reserves. We’ve finished a basic analysis of SA and will release results over the next few months. Low cost off-river PHES requires a large head and a steep hill (to keep the pipe short) and not too far services (roads, power lines etc) – all of which exist east of the gulf. PHES is less than a quarter of the cost of battery storage. Battery storage looks good behind the meter competing with retail prices of electricity. Regards, Andrew
REPLY to second email which asked for some clarifications. ….Qld weather differs from southern weather. We aim to publish this year, and have quite a lot more work to do. Separate from Mark Diesendorf. We are an ANU team. The results are not so surprising given the amazing reduction in PV and wind costs.

I now realize that Andrew Blakers is an ANU Professor, a heavy renewables promoter and funded by ARENA. (Also on the ARENA advisory board.)
Comments on Professor Blakers efforts will have to wait until he publishes. (I was unable to find out where his results will be published.) His article on ‘The Conversation’ shows that a man and his pea and thimble tricks won’t be disregarded lightly.

Like

The Achilles heel of nuclear is permanent disposable of used fuel. The Achilles heel of renewables is intermittency.

I guess you are bowing to a high priest in saying that. Between ourselves, I would say that the Achilles heel of carbon fuels is permanent disposal of their waste. The Achilles heel of renewables is baseload. Nuclear has long since solved both problems.

Like

Major Challenges in Renewable Integration, Report Says
Thomas Overton
2016 Sep 21
POWER

points to the issues and experiences which followers of Brave New Climate already recognize. Still, having a single review should be helpful, despite the optimism therein that the problems will be solved.

Like

To Roger Clifton.
Yes Roger, obviously. I have been a bit stronger in my language when contacting other anti-nukes and the reply has usually been no reply.
I wanted to catch a fish this time.
Regards

Like

Two commenters on the BNC forum mentioned Moltex Energy and their concept for a fast reactor. Some slides for a technical presentation are in a pdf entitled
The Simple Molten Salt Reactor Practical, safe and cheap.

I am encouraged by the fact that no reprocessing is to be done as part of the reactor. The actinide fluid is contained in slender rods which are removed after some time. Presumably some out-of-reactor reprocessing is possible.

Somewhere there is a price estimate of US $45/MWh for 1 GW units. While likely optimistic, it is less than half of the estimates for other fast reactor designs that I know about.

Like

The article referred to by DBB is at http://www.powermag.com/major-challenges-in-further-renewable-integration-report-says/?

It’s frustrating to note that NPP’s are now being recorded operating continuously and available for full load operation for 2.5 years or more, as against renewables, which cannot be relied on to reach 2.5 days, with the sole exception of hydro drawing from storage, which in most cases would make more money operating as peaking plant.

The undesirability of low levels of availability, reliability and capacity factor is lost on the majority of the population.

For any newbies present and typically calculated over a calendar year:

Availability Factor AF is the ratio of hours technically available to generate divided by hours passed. Typical figures for solar <50%, wind (?)90%, coal, nuclear and OCGT >90%

Reliability RF: There are various ways to calculate reliability, which is essentially similar to AF after allowance for planned maintenance has been made. The primary objective is to calculate the percentage of the time that the plant is available to generate at full load, divided by the total time that has passed minus the time lost due to planned maintenance outages.
Nuclear and fossil fuelled generators are typically 90+%. Weather-dependent renewables suffer due to the vagaries of the weather, although battery support and other energy storage systems can lift availability from less than 50% to perhaps 75%, but in these cases the cost of energy storage should be included in the capital and operating costs for the wind or solar power.

Capacity Factor CF is the amount of energy generated (MWh) in a period, divided by that which would have been generated had the plant generated at full load over the same period, which is typically 12 months. Typical PV 20%, Wind 30+%, Offshore wind 40+%, Fossil fuels 80+%, Nuclear operated as base load 90%, dropping to 75+% when load-following.

When I see AF and CF figures for weather-dependent renewables which have been calculated on the basis of a 1-day or less look-ahead instead of for the whole study period, my bullshit meter flies off scale. It is not uncommon for wind proponents to use this trick to try to avoid including the effects of low-wind weather patterns or for solar proponents to avoid the night hours and forecast clouds.

It can be difficult to identify the amount of energy generated which is sent to storage and which thus should be excluded from the figures quoted for energy produced. Energy sent to the customer is what counts.

Energy sent to storage is irrelevant from the user’s perspective. Only that portion which is reclaimed from storage to do useful work is ultimately marketable; 30% (CAES) to 80% PHES).

IMHO, the most important two are Capacity Factor, CF, and Energy Sent Out, ESO, although ESO is perhaps best expressed as a system-wide LCOE where storage and recovery are included in the computation of CF.

I raise these subjects because of the entirely irresponsible focus of media and anti-nuclear power advocates alike to focus only on Chernobyl and nuclear armament, neither of which is relevant to discussion of modern (Post-1980, say) power station designs.

Like

Are insurers offering cover for the risks associated with battery storage? Is this limited to specific battery types such as those in Samsung mobiles, or does it include others, eg lead-acid batteries, which are commonly required to be in fire resistance rated and acid-resistant rooms?

Of course, fire risks are not associated only with phones or battery panels. The volunteer fire brigade of which I am a member has attended several house fires which were started by overheated laptop computers, including one which originated in our Brigade Captain’s daughter’s bedroom.

Like

RE: Professor Andrew Blakers and pumped storage, he has been pushing this barrow for quite a while to overcome the limitations of renewable energy sources, so I do not envisage anything new coming from his forthcoming paper.

https://theconversation.com/how-pushing-water-uphill-can-solve-our-renewable-energy-issues-28196

As Robert Hinds has pointed out, his background and funding make him less than a disinterested party in the energy debate
https://theconversation.com/profiles/andrew-blakers-3328/articles

Like

SE asks, “Are insurers offering cover for the risks associated with battery storage?”

… including of course, cover for spillages of dissolved lead, to remove traces of lead from the neighbourhood soil, soil profile and water table to EPA standard. Localised pollution around an old power station can be made very expensive. By that standard, we should not let distributed generation become distributed pollution.

Like

Re old power stations, I had a pleasant surprise a few years back when entrusted with soil sampling and testing at depths from 0 to 70 metres in a total of over 3000 plan locations. My supervision team included three other engineers. Work was done by substantial, experienced contractors.

The two power stations were subsequently sold to AGL by the NSW government.

The results indicated very few major contaminated sites, mainly oil contamination to surface soils associated with oil storage and handling.

Lead, though tested for, was rarely encountered.

IMHO, the greatest risk is associated with fire, including electrical fires associated with switchgear and chargers for batteries of all descriptions. Plus, of course, those pesky Samsung phones.

As a past designer of structures such as schools, hospitals and the like, my guess is that insurers will soon realise where their primary risks have changed and put exclusions in standard contracts.

Like

6 Nuclear Energy Companies Building Molten Salt Reactors
Nanalize 2015 Oct 23
nanalize . com

briefly describes what each of these startups are doing. Moltex Energy still appears to have the best short range design.

Like

Re Peter L:
I am waiting for details of what failed and where, but unfortunately the only reports I have seen thus far have been so seriously polarised that at best they consider only a fraction of the story.

No doubt, small petrol generators are selling like hotcakes and will continue to do so for a very long time. There’s little more persuasive than a mother with a fridge full of rotting food and a couple of toddlers who don’t understand why none of the electrical equipment in the home works.

It’s always the folk at the end of the line who suffer the longest blackouts and the end of the (brown coal) transmission line is SA. Re-energising will progress west to east and south to north from the synchronised supplies, ie Victoria. This difficult-to avoid truth will also challenge the dreams of ANU’s pumped-hydro-plus-renewables-are-all-we-need Professor Andrew Blakers.

As Tasmanians discovered not so very long ago due to the failure of Basslink, duplication of essential HV transmission lines is not a luxury and the Heywood Interconnector and other transmission lines between generators, even wind generators and load centres are not exceptions to the rule.

A decent spinning reserve is similarly not a luxury. If it happens to be a GT then it also has black start capacity and thus the ability to re-start islanded portions of a damaged grid in which it is embedded.

How many batteries and inverters does it take to power a whole state for several days?

Who still wants essential services to be designed by politicians and populists instead of engineers?

Like

Thing is, the debate has started on The Conversation and after the dust has settled it would be great if Barry would put up a definitive article on BNC. http://theconversation.com/what-caused-south-australias-state-wide-blackout-66268

On 29 September 2016 at 13:55, Brave New Climate wrote:

> singletonengineer commented: “Re Peter L: I am waiting for details of what > failed and where, but unfortunately the only reports I have seen thus far > have been so seriously polarised that at best they consider only a fraction > of the story. No doubt, small petrol generators are selling” >

Like

singletonengineer,

Thank you. Would you like to interpret the Watt Clarity video for folks here (including me) and give us your insights? My first reaction was far too much wind and insufficient dispatchable power generating at the time.
First shot at trying to understand what went on, leading to the SA blackout of Wednesday 28th September

http://www.wattclarity.com.au/2016/09/first-shot-at-trying-to-understand-what-went-on-leading-to-the-sa-blackout-of-wednesday-28th-september/

Like

I didn’t find Paul’s video added anything to my own investigation on the event from the AEMO records and the Aneroid Energy site. The pricing analysis only cluttered up the issue.

Here’s my take on the event – I do not believe that the SA blackout was caused by wind turbines being shut down to avoid overspeeding. If look at yesterday’s wind generation at the Aneroid Energy site and isolate South Australia, you will see that all was going fine with strong but not excessive output of around 70% capacity factor until the grid suddenly collapsed
http://energy.anero.id.au/wind-energy/2016/september/28

If you also check do the same for fossil fuels (you need to remove some Qld power stations that aren’t eliminated by the state checkbox), you will see the real issue is the very low amount of fossil fuel generation that was taking place at the time. The Ladbroke OCGTs were flat out with Hallett ready to ramp for coping with fluctuations but the only other thermal generators online were the Torrens B units so there was very little synchronous inertia available to stabilise the grid.
http://energy.anero.id.au/fossil-energy/2016/september/28

The indications are that the Heywood interconnector was lost, islanding the South Australian grid creating a non-credible event. With the high amount of wind production and low amount of synchronous inertia, it wouldn’t have taken much of a trip event to cascade down the whole grid as there would not have been much inertia to drive current through the trip so it could be isolated.

Or it may be that the trip was caused by a lightning strike unloading the system, throwing some of the online generators out of synch (those smallest and closest to the strike would have been most affected). Once one or 2 of the generators were tripped, the demand would have tripped the rest.

The total lack of synchronous inertia of the doubly fed and full converter wind turbines that have been installed in Australia meant the South Australian grid was almost bound to blackout once the Heywood interconnector was lost with so few thermal generators online.

It all played out along the lines of one scenario from the following AEMO/ElectraNet report yet the AEMO states that it was wholly a weather event
https://www.aemo.com.au/-/media/Files/PDF/Joint-AEMO-ElectraNet-Report_19-February-2016.ashx

Like

EN, Greg and Peter,

I cannot at this stage add anything of significance.

Others on this site and elsewhere have pointed to the folly of relying on excesses of unreliables with insufficient synchronised capacity (rolling reserve) available.

The weather certainly was atrocious, but apparently no worse than 50+ years back. Are there records of state-wide blackouts from that event? I think not, otherwise they would have been cited.

My impression is that this week’s event is not entirely unprecedented, yet the resulting blackout, measured by extent (the State of South Australia), duration (Two days and counting, TBA), and cost ($Billions… who knows?) exceeds anything experienced previously.

Thus, despite whatever the politicians might say, we have evidence of a degraded system. A rigorous examination of the facts is in order and I have no doubt that AEMO and various others will provide their reports in due course.

South Australia might be positioned to give the whole world an education about the constraints that govern successful integration of high penetration unreliables in an existing system which, coincidentally, has virtually stagnant load growth year on year.

I used the word “govern” intentionally, because words and aspirations do not govern engineering systems: physical laws and engineering realities do.

In particular, I await the inevitable response from an ANU Professor who claims to be an engineer but who is not adequately qualified to be a member of the Institution of Engineers, Australia, one Andrew Blakers. Professor, this is your chance to explain again how 100% wind+solar, about half of which will be in Queensland in order to pretend to smooth the generation and demand curves, plus a few batteries and a modicum of pumped hydro can achieve that which was impossible in SA a few days back.

Professor Blakers, South Australians and the whole world await your analysis of the events of the past two days in South Australia. After all, you are a contributing architect to the substitution of Unreliables in lieu of the pre-existing and safer and more resilient electricity generation and distribution system in South Australia.

UNSW’s Prof Diesendorf might also provide valuable insight as to how, if his version of the 100% unreliables in a geographically distributed configuration would have avoided a similar failure, and at what cost.

Some might think that both will choose to remain quiet and thus escape critical appraisal by their peers and by representatives of the power generation industry in Australia, Germany, Spain, The Netherlands, Italy, USA, China, South Korea, Britain and other countries which have experimented with ever-increasing proportions of Unreliables, AKA Renewables in an attempt to lower carbon emissions to the atmosphere.

I make no secret of my reluctant recognition, several years back, that without nuclear power in the mix, we will fall short. Typing these words saddens me – many of us wished that it was not so, but an increasing number of thinking, numerate, educated Australians understand that there are no alternatives in a carbon-challenged world.

Like

Back to Mars.
Anyone watch Elon Musk the other day?

Even he (reluctantly) seems to think nuclear may be necessary on Mars. If only we had a way of convincing him of the need for nuclear on earth, we’d be home and hosed! Imagine Elon backing “Salt-X” or “Breeder-X”? This guy gets stuff done!

I’m loving the concept of a <$200 grand ticket for anyone who wants to go to Mars, but surely for the first 1000 people at least they would have to qualify by what the colony most needed at that stage? EG: I don’t know how many mail-clerks or artistic consultants they’ll need right away. Basically, I’m wondering how they’re going to do food in the first few years. How many tons of food would have to be shipped up there per person before the first colonials arrived?

Like

Thank you Greg Kaan and Singleton Engineer.

My expectation is that cause will be found to be very much as you both say. in a few words: too much asynchronous generation (wind and solar) and too little synchronous generation (coal, gas, oil, hydro, nuclear).

Nuclear could provide reliable power and supply 75% of our electricity, much cheaper and much more safely than renewables.

Like

Commotion at The Conversation.

In common with other armchair comments, this site seems to be concentrating on the small problem, which as a three-day storm that resulted in blackouts while being blind to the enormity of the recovery operation, which has been made much more painful, expensive and slow due to recent changes in the nature of the SA electricity generation and supply industry.

See: https://theconversation.com/turnbull-turns-south-australia-crisis-into-row-over-renewables-66297

A few familiar names have commented. A few more would not hurt.

Like

Re my post 2 comments upthread:

The Conversation’s moderators have removed my comment and many others as well, then closed the comments down.

What remains is a Bowdlerised remnant of its former self, devoid of all but the comments from the innumerate non-engineers, the politically correct and the sales reps for unreliable fanciful power systems.

My primary point was that restoring to service of SA’s current grid is much more complicated and time-consuming than was the case before the revolution,in part because it had to start at the Vic border and progress north and west from there. Formerly, restoration would have been possible radiating out from the synchronous, reliable, cold-start-capable coal fired generators, which SA no longer has any of. SA is learning the hard way that when they removed the old power stations, they needed to install at least a significant proportion of hydro, solar thermal, coal, biomass, nuclear or geothermal replacements. But they did not.

If the recent storms are the worst on 60 years, then why didn’t the BHP steel works at Whyalla close down then? Why not the lead/zinc refinery? Both have been crippled this time around.

The costs attributable to the elongated restoration process will be industrial, social, financial and political.

Maybe next time, South Australians will choose to have their power systems designed by engineers instead of spin doctors.

At present, SA’s collapse stands as a warning to the rest of the industrialised world that engineering requires engineers, not politicians, populists, salespersons and spin doctors.

Take a bow, South Australia!

Like

China leads new used fuel recycling project

China has operated two Canadian CANDU 6 reactors at Qinshan since 2003, and these have been used over the last few years to trial a new way of recycling used fuel from China’s main reactor fleet. In particular, uranium recovered from used PWR fuel is blended with a little depleted uranium to make natural uranium equivalent (NUE, about 0.7% U-235). This has been shown to behave the same as the natural uranium fuel normally used in those CANDU reactors.

This trial led to a 2012 agreement between Canada’s Candu Energy, China National Nuclear Corporation (CNNC) and two other Chinese companies to develop a detailed conceptual design of an Advanced Fuel CANDU Reactor (AFCR) based on the Enhanced CANDU 6 (EC6), which would run entirely on such fuel. One 700 MWe AFCR could be fully fuelled by the recycled uranium from four 1000 MWe PWRs’ used fuel. Hence deployment of AFCRs in China among its increasing fleet of PWRs would greatly reduce the task of managing used fuel and disposing of high-level wastes, as well as significantly reducing China’s fresh uranium requirements.

Now a new agreement among Candu Energy’s parent company SNC-Lavalin, CNNC and the major engineering company Shanghai Electric Group (SEC) has been signed, to set up a joint venture in mid 2017 to develop, market and build the AFCR. CNNC will have a majority share in the JV. Two design centres are envisaged, in China and Canada, to complete the AFCR technology, with a view to construction of two AFCR units in China.
WNN 23/9/16. China fuel cycle”
http://us1.campaign-archive2.com/?u=140c559a3b34d23ff7c6b48b9&id=4b25a0e8f9&e=a3b55276e6

I have a bit of a soft spot for the CANDUs and their derivatives.

Liked by 1 person

Moltex Energy sees UK, Canada SMR licensing as springboard to Asia
2016 Jun 28
Nuclear Energy Insider

lays out the Moltex Energy estimates for overnight capital costs, low, and an expectation of lower O&M costs than for light water reactor designs. For various reasons they don’t currently plan to attempt NRC licensing, which is too bad for the USA in my opinion.

Like

At first glance, it would seem to be an awful lot of effort to create an approximation to something they already have – natural uranium. Especially since the Candu’s are already rated for 1.2% fuel, which presumably will progressively replace the initial fuel as it depletes. However, the exercise is also providing an early customer for their nascent reprocessing industry. Eventually full-scale reprocessing of PWR fuel will provide the start-up fuel for the fast reactor fleet that will begin to overtake their PWR’s around 2050.

Like

I am interested to see that former police commissioner Gary Burns has been given the job of investigating the Grid failure that occurred in SA last week.

Does anyone know anything about him?

Google is fairly silent.

The site has me registered as Ton Carden but it is an error. It should be Tony Carden

Like

I think we should leave the SA storm alone. It’s bad PR, and will only confirm that us nukies are collaborating in an anti-sunnie, anti-windie agit-prop exercise. There was a huge storm that knocked out dozens of major HV powerlines. Even nukes would not have been supplying power that day.

Like

Innovative molten silicon-based energy storage system
2016 Oct 07
Science Daily

states the working temperature is 1400 °C. Is this too hot to connect to a supercritical carbon dioxide Brayton cycle turbine?

Like

Good question, DBB.

Published in 2004 and thus a little out of date, the following discusses theoretical operational limits of CO2 Brayton cycle turbines (Chapter 2) but cautions (Chapter 1) that advanced materials have not been developed for inlet temps above 650 to 700C (1200 – 1300F).

Click to access dostal.pdf

Sandia National Labs, 2013: More recently, (Slide 18) 650C is given as the practical limit, pending further advanced material development. http://www.osti.gov/scitech/servlets/purl/1221819 Future inlet temps of 850C are envisioned, but not for at least another couple of years (Slide 18 again).

Like

Not being an engineer I don’t know how to even ask this question, but if there’s a big lump of this molten-silicon material that can store thermal heat that high, would it be able to radiate heat or somehow exchange heat to another material that’s lower in temperature and actually can go through the turbine? Or could liquid tubes of XYZ go through the super-hot material and bring out enough heat to run a turbine?

Related: how many orders of magnitude cheaper does storage have to be before renewables actually can compete with nuclear?

Like

Too hot for turbine blades? Maybe, but 1400 C may be a useful top temperature for a magnetohydrodynamic (MHD) generator. With the input gas salted or raised to plasma temperatures, it becomes a conductor being forced through a magnetic field, thereby generating a current.

Nuclear heat generation doesnt have a theoretical top temperature — short of megakelvins — so the two would seem to be matched, however nuclear reactors have their own material limits.

https://en.wikipedia.org/wiki/Magnetohydrodynamic_generator

Like

eclipse now — The molten silicon-based thermal store will have process heat temperature. For example, aluminum melts at 1221 °C and some stainless steels melt at 1400 °C. The result for the heating fluid, transferring heat from the thermal store to the process vessel, is a lower temperature. If sufficiently low, depending on the process, the result might be working temperature for a supercritical carbon dioxide Brayton cycle. I doubt that this combination will ever find a commercial application, but one or two might actually be put into practice.

It is a bit embarrassing that the materials to fabricate a supercritical carbon dioxide Brayton cycle turbine with an upper temperature of 1400 °C do not exist at this time.

Like

eclipse now — A more traditional use of that process temperature thermal store would be just melting aluminum ingots for making castings. The working fluid brings almost 1400 °C to the ingot melt pot and is returned to the thermal store when its temperature falls to about 1225 °C to be reheated. That way everything is above the aluminum melting point.

Like

eclipse now — You asked how inexpensive does storage have to be so that intermittent generators can compete with nuclear generators. The answer is that it depends so I’ll just work out an example based on the situation in the Pacific Northwest. The dollars are US currency. Bonneville Power Administration states their wholesale price is $31.50/MWh; this is to recharge the batteries. With great regularity there are 3 weeks every fall with almost no generation from the 8,000 MW of wind farms here. There is less than full generation at other times so I assume that the required 8,000 MW of batteries, enough to cover the load for those 3 weeks in the fall, has a capacity factor of 0.15.

I assume that the competition is Moltex Energy small molten salt fast reactors, selling for a believable $42/MWh. Note that this is competitive on the Mid-Columbia hub and that it is less than half the price for new build light water reactors in the West; I don’t understand the Russian or the Chinese pricing systems. In any case, the batteries have to sell power for $41.50 which means all expenses, mostly capital plus interest, have to be met by the margin of $2.125/MWh between the sale price and the purchase price, less 20% losses.

There are 8766 hours per annum and 15% of those, for the capacity factor, is 1315 hours. So the income to pay for all the batteries and other expenses is $2794 per annum. This has to buy 576,000 MWH of batteries to cover the 3 windless weeks.

Somehow I doubt such will ever come to pass…

Like

Oops! I forgot to multiply by 8,000.
$2794×8000=22,352,000 per annum.

But if no interest and a 30 year life, that’s a total of $670,560,000 towards the batteries, etc. So $1172.31/MWh of battery. So if I haven’t made a mistake it looks possible now.

Like

Oops! I forgot to multiply by 8,000.
$2794×8000=22,352,000 per annum.

But if no interest and a 30 year life, that’s a total of $670,560,000 towards the batteries, etc. So $1172.31/MWh of battery. So if I haven’t made a mistake it looks possible now.

David, I’m afraid you are out by a factor of 1,000. You need 4,000 GWh (8GW x 21 days x 24 hours). Ex-EV battery packs might cost you around $100 / kWh right now (excluding inverters, charging, environment). The total cost would be 4,000 x 1,000,000 * 100 = $400bn.

Battery prices are never going to be low enough for 3 weeks of backup. They are only economic for a few hours. You need another solution for 3 weeks.

Like

Mars question again: is there a calculator that reduces the radiation by the amount of atmosphere? I know Mars doesn’t have a magnetic field, but there’s all this talk of manufacturing super-greenhouse gases that are 17,000 times more potent than CO2 to cook Mars up to comfortable temperatures. Then as the atmosphere cooks up and the CO2 at the poles melts, the atmosphere will increase. Now if we increase it to half an atmosphere or even 1 atmosphere pressure, and it becomes a whole lot warmer, people could walk around without pressurised space suits and basically just wear breather masks. But what about radiation? How much would the atmosphere reduce solar radiation without a magnetic field?

Like

Peter Davies — Thank you!

The original question was when would the intermittent generators be inexpensive enough to replace nuclear power plants. While I don’t want to make embarrassing errors, I am just pleased you are the one who noticed the mistake.

Eclipse Now has, finally, an answer to his question.

Like

Eclipse Now:

These values are from a bar graph with a logarithmic scale that I copied from some website over a year ago.

Annual Cosmic Radiation (Sea Level) 0.3 milliSieverts
US Annual Average, All Sources 4 mSv
Abdominal CT Scan 8 mSv
DOE Radiation Worker Annual Limit 20 mSv
6 Months on ISS (average) 80 mSv
180 day Transit to Mars 300 mSv
500 days on Mars 300 mSv

From this you can see that the Earth’s magnetic field ( and maybe having the earth blocking radiation from one side) cuts radiation dose on the ISS compared to interplanetary space. However, you can see that the mass of the atmosphere does most of the shielding for the surface of the earth.

The value for the stay on Mars must ignore the effect of piling dirt on the habitat.

Since the gravity of Mars is .4 of Earth gravity it would take 2.5 times the mass per area of atmosphere to give 1 atmosphere pressure on Mars. So the radiation shielding of a thickened martian atmosphere would be plenty.

Like

Hi Jim,
thanks for that. Here’s a reddit thread I started on the topic, and the discussion there was quite interesting.

“As long as the air pressure at the surface post-terraforming is at least 0.2 atmospheres, it will provide adequate radiation shielding (over 5000 kg per square meter column density). This also happens to be about the minimum air pressure that would allow us to walk around outside without pressure suits, just wearing oxygen masks so we can breathe. And there’s believed to be enough carbon dioxide frozen on Mars to make the atmosphere at least that thick.
So despite the popular misconception, a magnetic field is not at all necessary to protect the surface of a terraformed Mars from radiation.”

https://en.wikipedia.org/wiki/Health_threat_from_cosmic_rays#Shielding

“a NASA design study for an ambitious large spacestation envisioned 4 metric tons per square meter of shielding to drop radiation exposure to 2.5 mSv annually (± a factor of 2 uncertainty), less than the tens of millisieverts or more in some populated high natural background radiation areas on Earth”

So 4 metric tons of shielding per square meter is enough. Earth’s atmosphere has a column density of 10 metric tons per square meter, so we just need 40% of that. At 1 g, 40% of the column density would have a pressure of 0.4 atmospheres, but Mars has a surface gravity of 0.38 g so the air pressure would be 0.4 * 0.38 = 0.15 atmospheres. So 0.2 atm should be more than enough for shielding, and it’s just enough for the 0.2 atm partial pressure of oxygen we breathe.

Like

“The Department of Energy’s National Energy Technology Laboratory (NETL) will award up to $80 million to a 10-MWe pilot project that seeks to advance the development and commercialization of supercritical carbon dioxide (sCO2) Brayton power cycles.”

Not much more detail than the three research entities that will be involved with the pilot plant.

http://www.powermag.com/supercritical-co2-brayton-power-cycle-pilot-plant-bolstered-with-80m-in-federal-funding/

Like

Crescent Dunes has managed 5 days of continuous output in July!
They throttled output down to about 60% for 8 hours of the day but it looks like the storage was on its last legs in the hour before capturing resumed. See page 3

http://www.powermag.com/crescent-dunes-24-hours-on-the-sun/

Unfortunately, the period this trial took place was not given so we cannot check if there was any significant cloud cover during the trial.

It will be interesting to see if Solar Reserve can bring down the costs as they claim they will for South Africa,

Like

Thanks for that, Greg, but the Crescent Dunes article leaves more questions than answers.

Not a word about capacity factor, although from the graph it seems that average energy sent out on a hypothetical day is about 70MW, which for a 110MW plant represents only 64%. NB this small graph contains no data – it is a statement of expectation.

That assumes energy harvesting via the heliostats, for much of the sunlight hours, of 200MW.

That suggests that the field of mirrors needs to be increased by 50% to enable 110 MW sent out on the best summer day, in which case every other month apart from July will still achieve less than nameplate capacity. It says nothing about achievable performance in the winter months – perhaps a quarter of the above 64%? Or of the annual capacity factor – midway, at 50%, minus downtime for fires and maintenance?

This is typical for the magazine which you cited, which rarely provides other than partial, cherrypicked information and sweeping conclusions, which in this case end with:
Said Smith, “We’ve convinced the financial markets.”

The Powermag article did nothing of the sort. The markets will need comprehensive, year-long operational data, life expectancy and capital and operating costs, none of which have been disclosed as far as I know.

Real financial markets would be more interested in the $760M (US) government loan, the expected return for private investors and the guaranteed income, 0.135 US cents per kWh, which is several times higher than wholesale electricity rates in both USA and Australia.

Until these questions have been answered, CST remains a vanity project for rich companies and governments which want to be seen to be doing something for political purposes, regardless of the cost.

Like

singletonengineer — A small correction. The US DoE provides loan guarantees, not the loans themselves. The constructing company pays a six digit fee for the loan guarantee.

Like

On ABC Australia News Political Editor Chris Uhlmann, 2016 Oct 19, writes “South Australia’s storm caused transmission faults, but that is not the whole story”. It appears that he has access to the latest AEMO analysis which blames the frequency cutoffs used in the wind turbines.

I would go further in suggesting adding batteries for frequency control and also converting permanently idled generators into frequency stability units. There is a company in the USA which specializes in doing such conversions for retiring coal burners. As South Australia has some of those that might be a plan.

Like

Here is the extended AEMO report. The wind farms that tripped prior to the Heywood interconnector have been recognised to have overly conservative fault ride through settings and are being updated for greater tolerance to line faults. How this will affect the turbines in the long term remains to be seen but I would have thought that the tolerance levels were set according to the capability of the wind turbines to survive voltage fluctuations.

The most contentious portion of the report is likely to be the statement

Investigations to date indicate that information on the control system involved and its settings was not included in the models of wind turbine operation provided to AEMO during NEM registration processes prior to connection of the wind farms.

A local renewables proponent website has been aggressively putting the blame on the AEMO for not having the foresight to have obtain the ride through capabilities of the wind farms. My interpretation is that the wind farm operators failed to provided pertinent information on the limitations of their facilities during registration with the AEMO to join the NEM.

I foresee court cases for damages before this is all over,

https://www.aemo.com.au/Media-Centre/-/media/9027D5FB69294D408E4089249F38A36D.ashx

Synchrnous condensors would not have been enough to cover the shortfall left by the wind farm trips. Large enough batteries may have allowed time for large load shedding but what would be large enough? 500MW was lost before the Heywood interconnector tripped but how long do the grid operators need to shed that much load? I feel the amount required is not likely to be economically realistic.

Like

More big claims…

The SunShot Initiative supports research and development of concentrating solar power (CSP) technologies that reduce the cost of solar energy. CSP helps to achieve the SunShot Initiative cost targets with systems that can supply solar power on demand, even when there is no sunlight, through the use of thermal storage. Since SunShot’s inception, the levelized cost of electricity for CSP has decreased about 36 percent, from $0.21 cents per kilowatt hour to $0.13 cents per kilowatt hour, already over half of the way toward achieving the SunShot goal of $0.06 per kilowatt hour.
http://www.nextbigfuture.com/2016/10/argonne-labs-makes-solar-thermal-energy.html

Like

Greg Kaan — Batteries used for frequency control are typically at 10–15% of full charge and so have some ability to store excess generation giving the operator time to restore the balance between supply and load. In South Australia as soon as the transmission lines tripped there was excess supply. An appropriate setup might have kept the wind farms from all tripping off together, creating an excess demand situation in which batteries alone can do nothing. It would take some additional sensing gear; a local manufacturer is SEL, Schweitzer Engineering Laboratory, but other companies make similar protection devices.

Like

Sorry David but the event sequence from the preliminary AEMO report below clearly shows that there were no significant periods where there was excess generation. The 2 line faults preceding the wind farms tripping were single phase to ground shorts causing voltage dips which triggered the wind farm protection circuits. These created a generation shortfall which led to the Heywood Interconnector overloading and tripping, all in a period of 7 seconds.

Perhaps batteries could have filtered these voltage dips since they were of short duration but they would then be operating as line filters rather than general frequency support services.

https://www.aemo.com.au/Media-Centre/-/media/BE174B1732CB4B3ABB74BD507664B270.ashx

Like

I’m sure everyone realizes that the subsidies for nuclear in the U.S. (every country is different) is controversial as to what those subsidies are. The EIA uses a specific set of ‘narrow’ criteria that has been criticized by anti-nukes for not including all the externalities such as mining for ore, budgets for the National Labs and decommissioning, etc. Additionally it’s criticized for not including “military” spending and R&D. The former is legit IMHO but the latter is not since commercial nuclear energy’s relationship is an ‘after the fact’ (the antis say you have to include the Manhatten Project…in fact realy we are talking about the US Navy propulsion system. But then the Boeing 707 was a direct application of B-52 technology…where does this ever end??).

I think it’s important to broaden out the subsidy issue a bit anyway. But what IS important is including the ‘dividing’ of the subsidies over the lifetime of a project as some have alluded too. Thus 20/25 years for wind/solar, 60 years for nuclear.

Like

@Peter Lang’s “U.S. Energy Subsidies” graph at the top of this thread. The fine print on page 21 of the linked EIA reference informs:
“The credit for the production from advanced nuclear power facilities had no value in FY 2010 or FY 2013 as this credit does not go into effect until qualifying new nuclear power plants produce electricity.

Those “qualifying new nuclear power plants” would be the four AP1000 currently under construction at VC Summer and Point Vogle. They won’t produce electricity until late 2019 and 2020.

Only the first 6 GW new qualifying construction will receive the $2.10/MWh subsidy, which will cover these four 1150 MW plants plus the NuScale SMR’s currently being sited at INL, which might complete 2024.

I’m uncertain uranium mining is currently subsidised. It certainly was during the go-glow years of the 1950’s and 60’s. But production peaked at 16,800 ton in 1980, declining precipitously to 5,700 tons four years later. By 2003 domestic mines provided but 5% of the fuel consumed by US commercial reactors, when about 50% was being sourced from Russia in the “Weapons to Megawatts” program.

Domestic production has ticked up a bit since then, but I’ll hazard we’re still buying 90% of our new uranium from Canada and Oz, neither of which noted for freebies to the U.S. Domestic producers have contested the Administrations current plan to dump surplus defence uranium on the commercial market at less than cost and market value. That would constitute subsidy,

http://world-nuclear.org/information-library/country-profiles/countries-t-z/usa-nuclear-fuel-cycle.aspx

Like

Eclipse Now — It appears that GE-Hitachi and Southern Nuclear are going to try to obtain a grant from the DOE for advanced nuclear reactor research. I have no idea whether this will lead to an attempt to commercialize the PRISM in the USA.

Like

It may just be a good time to suggest reviving some plutonium-burning designs, such as PRISM. The US is currently under pressure to show a track record of burning its surplus military plutonium, following Russia’s repudiation of the US ditching its commitment to burn it as MOX.

The PRISM falls short of the IFR concept because it does not “integrate” with on-site reprocessing. In the current instance, that’s a real selling point. Instead of not needing refuelling until the cows come home, the PRISM needs regular refuelling with the unloved plutonium and produces a steady stream of hard-burnt used fuel useless to any would-be weaponiser. Burning 35 tonnes of plutonium would take a lot more than one (300 MW) PRISM, but it would be seen as a good start.

Like

With regard to Andrew Blakers and Mark Diesendorf, I’m curious about how much outside mainstream scientific view their opinion is. The 30 or so Australian scientists who have signed the Brook Bradshaw letter have an average h-index of about 50. I have been unable to find a single scientist with an h-index > 50 in Australia who supports the 100% renewable scenario. Diesendorf’s h-index is 22 according to Google Scholar, whereas Blakers is about 23 and Ian Plimer’s 22.

9 of the top 10 conservation Biologists in Australia can reasonably be said to have publicly supported nuclear energy.

I have made the statement on several other sites that “no Australian scientist with an h-index > 25 opposes nuclear energy and scores support it.” Thus far no one has been able to provide a counter-example.

Surely we do in fact have a 97% consensus that nuclear energy is necessary to address climate-change. Is this testable? Would there be any joy in having a parliamentary committee survey the scientists? I spend much more time arguing with anti-nukes than I do with climate-change-denialists. Surely the anti-nukes are by far the greatest risk to humanity

Like

Bill Schutt — I would rather say that nuclear power is highly desirable, but possibly not necessary, in changing to a low greenhouse gas emissions world. The question is cost and environmental factors. On both nuclear power scores well despite the appearance that intermittent generators do well on cost provided all the costs and risks are not considered.

Power planners usually want a variety of generator types to hedge against contigencies; the recent experience in South Australia suggests why.

Like

Yes, David, possibly not necessary is about right. No one can exclude a disruptive breakthrough. To argue with the doctrinal anti-nukes, you only have to believe that 100% renewable energy is a less than 100% certainty. So suppose one were to conclude that it was a 50:50 split, then clearly the safest option is to have both nuclear and renewable energy. But my guess is that there is very likely a 97% or similar consensus amongst our most eminent scientists that nuclear energy is necessary. But we need to do something to get people to accept that going for 100% renewable energy has a very good chance of causing a climate catastrophe.

Like

Bill Schutt — A climate catastrophe we are going to have unless, possibly, drastic action is taken to lower the carbon dioxide concentration well below the current 400 ppm. For from the Wikipedia article on Pliocene climate we learn that the sea level will equilibriate at around 25 meters higher than now, that being the approximate value for the mid-Pliocene, about 3 million years ago and the last time that the carbon dioxide concentration was as high as now. As the Isthmus of Panama closed well before then, the ocean circulation was much the same as now and so the climate system was as well.

Eminent scientists rarely have the right training and experience to provide expert witnesses regarding power system design. I know one, from undergraduate days, who has recently been awarded 3 big prizes and is in line, potentially, for a share of the Nobel Prize in Physics. If asked about this, he would say that he had no expertise to offer and decline to express an opinion.

I point out that Norway is 98% renewable power, almost all hydro. Likely when the coal burners are at end of life the units won’t be replaced so Norway could then claim 100% renewables. But we understand that Norway has rather special geography to share among only around 5 million people.

Like

Vietnam Scraps Plan for Its First Nuclear-Power Plants
Vu Trong Khanh
2016 Nov 10
The Wall Street Journal

so coal burning will rise from the current 30% to 55% over the next decade, according to the article.

Not in the right direction. Possibly with SMRs available the Vietnamese planners will avoid any other coal burners.

Like

Nuscale sees U.S. Market for Modular Reactors, Others Don’t
Bloomberg BNA
2016 Nov 07

is a good presentation of the various points of view regarding the Nuscale LWR SMR, together with a brief summary of the other vendors efforts for other countries. Nuscale continues to state that the design will only cost US $5 per gross watt generated, a good price these days in the USA, but not as inexpensive as the South Koreans in the UAE.
Nuscale now states that construction will only require 36 months once the civil works are complete. That is comparable to the construction time for a combined cycle gas turbine.

Like

The article seemed to me rather unsympathetic. The author’s idea of balance is to quote someone from the Union of Concerned Scientists saying that no matter how small it is, it is still nuclear and therefore dangerous. (Those guys aren’t scientists!) It also quotes an economist saying that gas will always be too cheap for nuclear to compete. There was no mention of a future carbon price.

There also seems to be errors in describing the NuScale concept. The author implies that the thing is chopped up into three chunks before leaving the factory, for separate delivery. However the NuScale website says, “The small size of the completed module permits [completed] component shipment via conventional large object transport such as truck, rail, and barge.” That is, it is a module, a single module to be delivered in one piece. (image)

The article goes on to imply that the module is on site for 36 months before being useful. Presumably most of that time is spent with local people trying to weld the three chunks back together. On the contrary, the NuScale website refers to a critical path of 28.5 months. Compared to gigawatt reactors, the SMRs offer the major selling point of each module being able to start paying off its cost within about two years, while the rest of the power station grows more modules.

Curiously the article says that the NuScale only carries 5% of the fuel load of the gigawatt PWR. I think the author is confusing the mass of the fuel with its enrichment level of 4.95%, the same level as most gigawatt PWRs.

Like

Roger Clifton — Thanks. I assume that the three chunks are the nuclear module, the steam turbine and the synchronous generator. The latter two bolt together and the first two require some pipefitting. Oh yes, there is also the condenser and the small cooling tower. More pipefitting.

Like

For a system advertised as plug-and-play, three years (or even 28.5 months) seems to be an awfully long time between the module arriving on site and its first electricity production. Still, experience has probably taught them to leave plenty of contingency time in the critical path analysis.

Perhaps that is just for the first module. Since the team gets to do the entire operation 12 times during the growth of the 600 MW power station, there is plenty of scope for learning. The last module might be installed in a fraction of that time. Their next power station in a city not too far away would benefit from the learning achieved by the team.

Like

Do we know yet whether this is the construction stage for the first module? If so, there will be the usual Balance of Plant (BOP) items that include station plant. Not all need to be repeated for modules 2 to 10.
I am not familiar with the proposed site and plant layouts, but for example, if a chimney stack is needed to vent spent gas, or cooling towers are needed for spent steam, they might take a longer time but only one or two might be needed. Duplication of shared plant is advisable in order to permit maintenance without taking all units out of service.

Similarly, water supply and water treatment for the secondary feed water system which needs demineralised water in order to avoid poisoning the turbine.

Hydrogen storage and/or generation plant for generator cooling?

Compressed air?

Station power supplies, control room(s), administration building, stores building, sealed roads and carparks, security fencing and systems, staff training facilities (plus simulator?), communication, lubricating oil storage facility, switchyard and transformer yard for stepping up the voltage from the generator’s output to match that of the HV transmission line to the nearby city/load.

Indeed, I’d be very surprised if the station plant and first unit of a 6 x 60 = 600 MW facility could be constructed within 28.5 months, but possibly it could with prefabricated buildings.

On the other hand, I agree that 28.5 months is a longish time between units. A couple of decades back a set of 4 x 660 MW coal fired units were commissioned at intervals of 6 to 9 months at Bayswater Power Station.

So, a nest of 10 or 12 x 60MW bolt-together SMR’s could perhaps commission the first unit at the end of (guessing!) Year 3 or 4, but the remainder might be completed at 4 month intervals = project completion 6 or 7 years from turning the first sod.

Of course, if a brownfield site is chosen, then the former coal fired plant would have some of these services, but in the real world, the inner perimeter of the site would probably need to be cleared and started from bare earth.

Further discussion is probably not advisable unless with a construction schedule in front of us covering design, procurement, manufacture, construction, commissioning and licencing. What is the plural of “licencing”? I imagine that licences will be stepwise through at least four or five primary stages, for the plant as a whole and for individual units.

A further question: Will the turbines of such a plant be shared plant, say 180MWe, or individual tiny 60MW items served by individual reactors?

Let’s not base our plans on guesswork, or we will wind up looking very silly… and disappointed.

Only promise that which we know can be delivered. Unreliable wind and solar aficionados are still learning that lesson the hard way.

Like

Roger Clifton & singletonengineer — Each module has its own turbine and generator as well as its own condenser and evaporator, according to various artist drawings on the Nuscale website. Shared services have to be constructed first, principally the large excavation for the 12 modules. Once that is complete, onsite construction of the modules can, in principle, be done in as much parallelism as the reactor module factory can manufacture the units.

As for 28.5 months to mechanical completion, that is fast. A large combined cycle gas turbine takes about 3 years after site preparation is complete.

Like

Thanks for the reminder about the configuration, DBB. Since we are discussing generators of only 60MW, we can probably forget hydrogen cooling as well.

This points to a major stumbling block for nuclear power, which is the slow build rates.

The entire mechanical and electrical construction and commissioning time for the coal-fired Bayswater Power Station was 1980 – mid-1985. 5.5 years for 2640MW, ie 480MW/year on a single site, with civil and site works starting a couple of years in advance. That is the equivalent of one NuScale every month or two.

30 years ago, new coal fired power stations were planned and constructed in timescales of years. The timeline for new nuclear is currently measured in decades and growing longer.

It seems that the primary difference between the two is safety: Coal is far more damaging than nuclear, yet is subject to astonishingly detailed and expensive timelines primarily due to the regulator’s approach to micromanagement of safety related issues far beyond anything that other industries are subjected to.

Consequently, an AP1000 constructed in USA costs twice as much and takes twice as long to construct as its South Korean sister.

The present system simply isn’t delivering what is needed and the competition is heating up.

Indeed, the transition away from the American regulatory model for safe design, construction and operation of NPP’s is probably so well advanced that the future is clear. South Korea, China, France, Great Britain and others are in front. Additionally, NuScale’s majority owner, Fluor, is well established in China and South Korea.

For an aggressive, pessimistic overview of the role of the DOE and NRC in relation to NuScale and the prospects of nuclear power in USA, read this article in Fuel Cycle Week dated June 2013:

Click to access FCW-525-6.13.13.pdf

It is over 3 years old, but nothing has changed since.

Like

The two Westinghouse AP1000 nuclear power plants under construction at the V.C. Summer site are taking 6+ years to construct, each, and are on schedule to cost US $5.87/W. The South Korean reactors in the UAE are on schedule to be completed in about 4–5 years at a cost of around US $4.5/W, depending upon the exchange rate.

Like

What does 5.1 $/kW pay for?

Page 22 of the NuScale presentation, shows eight categories, including site assembly and balance of plant.

The units are missing from the table, but they turn out to be millions of 2014 US dollars per 570 MW of capacity, hence $5078 per kilowatt. Curiously, this is the cost for a FOAK, so presumably the cost would come down later for a NOAK.

Like

Not so very long ago I read $2,500/W capacity for S Korea and $5,000 for USA. Can’t recall where. If I find it I will post accordingly.

Like

That same presentation (here, p 26) elucidates the timeline:

“Schedule based on 51 months mobilization to mechanical completion.
28.5 month critical path – first safety concrete to mechanical completion.”

That is not 28 months per module but 28 months for the entire power station of 12 modules and 570 MW. This (the financier’s?) clock starts after all the earthworks and concrete have been completed, and ends at “mechanical completion”, after which pressure tests, logistics and power ramping still have to be completed.

It is not clear (to me) whether “mobilisation” starts when the NuScale factory starts on the first module, or when the first module is taken off-the-shelf at NuScale and placed on wheels, or even when the first bulldozer enters the site. How much time that the bridging loan must pay high interest does depend on when the first, and then the last cash hits the till at NuScale. Converting the capital would become cheaper once the station starts selling power. Impressive as 28 months sounds, the longer timeline of 51 months warns us that the full commitment is still longer than four years.

Like

Costs of anything whatsoever, varies with place, time and circumstances. The Indian prototype fast reactor, nearly complete now, is estimated to cost 843 million dollars or $1700/MW. Both fast reactors and prototype are cost increase factors.
https://en.wikipedia.org/wiki/Prototype_Fast_Breeder_Reactor
I wish they also have a version designed for 20%LEU fuel. 20% LEU could be imported but plutonium, even reactor grade, is not traded.

Like

Hi Jagdish. A FOAK costing of only $1700/kW of capacity is surprisingly cheap. The first unit (first criticality next year) is to be followed by seven production versions of the same 500 MW fast reactor before eventual production of 1000 MW reactors. The world will be interested in how costings for the NOAKs prove out.

Production versions of fast neutron breeders will be crucial not just to India, but to an eventual global rollout of nuclear electricity, once the world wakes to an emergency and mobilises the only feasible large-scale energy source that can completely replace fossil carbon. However the timescale of the rollout is dependent on the doubling time (pages 8+) to fuel the expanding fleet. Oxide MOX currently in the PFBR does not double fast, but the PFBR could be converted to metallic fuel, which doubles faster.

Like

https://www.greentechmedia.com/articles/read/comed-strips-mandatory-demand-charges-from-illinois-energy-bill?utm_source=Solar&utm_medium=Newsletter&utm_campaign=GTMSolar

It looks like Illinois may now pass a bill acceptable to most parties which will allow the Exelon Quad Cities and Clinton nuclear power plants to stay open at a cost of up to $265 million per year. Apparently they have lost around $800 million over the past seven years.

There is a sweetener for a couple of Dynergy coal-fired plants involved but the environmental lobby are now supporting the bill. Previous versions had various show-stopper clauses.

Someone here was very concerned about both nuclear plants being shut over the next 18 months, but I forget who.

Like

Click to access A-71-46_e_V1604696.pdf

Clause 46 and 47 appear to say that the coal power cycle is responsible for more than half of the total collective dose to the local and regional public from the discharges due to a single year’s global electricity generation.

It also ranks PV and wind above nuclear power, due to the impacts of rare earth mining and milling.

Further, since almost half of the impact of nuclear power is due to mining and milling, how sensible is it to not consider nuclear power in a country or state where uranium mining is legal and socially acceptable?

This paper appears to rank, from a public and employee health perspective, electricity generation options as (worst first) Coal, Solar PV, Wind, Nuclear, Other.

Geothermal might also rank above nuclear, depending on radon emissions.

Thus, rational planning of zero CO2 electrical power boils down primarily to matters such as cost, availability, reliability and scaleability and less on health-related factors.

Those who seek to place health at the top of the list of reasons for excluding/avoiding nuclear power must first demonstrate where the UNSCEAR report is wrong – and UNSCEAR, which represents almost 30 countries and is very much cross-disciplinary and global in its approach has been reporting on this subject since the mid-1950’s.

Like

See also: http://mzconsultinginc.com/?p=846&print=print

Note especially:
“…in all cases these levels of exposure are relatively low and have little impact to public health…

“So why talk about this? The reality is that this information is not likely to change even one single mind on whether someone supports nuclear power or fears it. We live in a world where facts no longer matter – the only truth is the one that any one person believes. Well, we believe that scientific study remains the best way forward to establish truth and that studies such as these are part of the path forward.”

Meanwhile, news of massive coral reef bleaching events arrives daily and the root causes of climate change, in all of its forms, go unaddressed.

Like

UChicago startup turns renewable energy into natural gas
Greg Boro
2016 Dec 01
Phys . org

Electricity splits water and the resulting hydrogen is combined with carbon dioxide to form methane via an appropriate microorganism. The carbon dioxide might be that in biogas, about half carbon dioxide and half methane, from waste water treatment.

This looks quite feasible at the industrial scale, but requires a good source of carbon dioxide as well as the required ‘excess’ electricity. Maybe almost every municipal waste water treatment plant will have one.

Like

“Maybe almost every municipal waste water treatment plant will have one”

Methinks DBB speaks with tongue in cheek. For many years it has been standard practice for sewage plants to feed their fermentation gas into a standing diesel engine, generating stable power to go back into the grid, which is much more useful than methane of dubious quality.

The University of Chicago could have titled the project “Turning electricity that no one wants into an undesirable greenhouse gas”, but I guess that wouldn’t attract giddy young students to a fashionable cause. There must be better ways to fund and staff University laboratories to do Good Things, even if the projects don’t impress in campus coffee shops.

Like

Roger Clifton — The waste water treatment biogas is often burnt in an onsite gas turbine. I doubt that biogas will run a diesel engine. In case I am wrong about that, please provide at least one reference.

The result of simply burning the biogas, of course, is that the carbon dioxide component is just passed through to the atmosphere. The activity described in the article produces almost pure methane, good enough to put into the natural gas pipes. Obviously, this is done with the excess not required for energizing the waste water works.

Some waste water treatment plants produce more biogas than is locally required for operations. For example, the San Diego waste water treatment plant produces enough extra to make it worthwhile to refine out the methane. The result is of sufficiently high quality that Linde Corporation buys it to resell as bottled gas with the excess trucked to the nearest entry point for the natural gas pipelines.

As the article points out, there is a pilot operation with a small industrial scale plant either still building or in operation already.

I don’t make reference, usually, to all the laboratory results, most of which are unlikely to scale. This one does scale.

Like

I stand corrected. Small generators running on natural gas normally require an ignition system and a lower combustion ratio, so it is more likely that a gasoline engine is converted to methane fuel. Even so, diesel engines can be adapted, at some cost: (ref:dual fuel) and (fuel conversion). My hazy memories from the distant past were only of an oily standing engine thudding away in a shed.

Yes, I really did think you were joking. We are surrounded by people who imagine that they are saving the greenhouse by replacing thermal generators with windmills backed up by methane-leaky gas turbines. Considering the extra threat that methane pipelines imply, it seemed to me that a system that converts sanctified (wind or solar) energy into yet another source of methane, was a grim joke to be shared among the few of us.

Like

Leave a Reply (Markdown is enabled)

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s